[go: up one dir, main page]

WO1996030844A1 - Systeme de support destine a l'automatisation de la preparation de structures d'histoires - Google Patents

Systeme de support destine a l'automatisation de la preparation de structures d'histoires Download PDF

Info

Publication number
WO1996030844A1
WO1996030844A1 PCT/JP1996/000813 JP9600813W WO9630844A1 WO 1996030844 A1 WO1996030844 A1 WO 1996030844A1 JP 9600813 W JP9600813 W JP 9600813W WO 9630844 A1 WO9630844 A1 WO 9630844A1
Authority
WO
WIPO (PCT)
Prior art keywords
story
narrative
agt
event
tree
Prior art date
Application number
PCT/JP1996/000813
Other languages
English (en)
Japanese (ja)
Inventor
Takashi Ogata
Original Assignee
Takashi Ogata
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Takashi Ogata filed Critical Takashi Ogata
Publication of WO1996030844A1 publication Critical patent/WO1996030844A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/40—Processing or translation of natural language
    • G06F40/55—Rule-based translation
    • G06F40/56—Natural language generation

Definitions

  • the present invention improves the human interface capability of a computer based on a mechanism for automatically creating a story outline (art line) more efficiently and diversified than a human being, thereby improving human creativity. More specifically is to M to a system that is intended to support the idea specific working process c, the narrative structure created automated support system, descriptions of fractional one or more events (out Raikoto) And inputting the setting of multiple parameters that mean the request from the user (input by the user), referred to as core stories, stories, pros, professionals, without the need for human intervention. In three stages, the description of the conceptual structure (called ⁇ story structure j), which is the outline of the story, is automatically created.
  • ⁇ story structure j which is the outline of the story
  • a story means a form of discourse composed of an organic combination of a number of events.
  • the story structure means the data in the description format inside the computer system of the story.
  • the present invention is based on the background described above, and considers the extent of human creativity and the sexual factors.
  • Communication iSK and I Cormorant is intended to contribute to the realization of the major goals r
  • the present invention is, usually to the culture of soft toe in the culture of the people ⁇ ;.. Focusing on the story should be called r Ma, this story ability to create computer software Touea By providing a fundamental method for realizing the goal.
  • r Ma this story ability to create computer software Touea
  • the distinctive functions of narrative and human narrative generation are the aesthetic expression function that evokes catharsis or excitement.
  • the semantic generation function through the arrangement of fragmentary thoughts, the concrete expression of abstract thoughts Emotional expression function for transmitting knowledge.
  • a basic system for effectively utilizing and realizing these characteristics of the material is provided as one of the above-mentioned bet requirements for the future machine technology. We thought that it could be used as nuclear technology.
  • natural language interface, marriage meti-me interface, gaming & simulation, intellectual education system decision support 3 ⁇ 4 support creation of thousands of skulls. Support for creation of scenarios and advertisements, ideas and embroidery.
  • the present invention provides a narrative structure that can support creative activities and can be applied to various fields of application such as Humanintuff X-S, in view of the current state of the multimedia age.
  • This is a W to create automation support system.
  • the current state of computer software including the latest technologies such as Multimetia, is that software that can achieve the specific goals (in this case, narrative structure) that human soft-to-X creators want to create. Designed and described one by one, this software cannot play any role other than the operation originally assumed.
  • the present invention realizes the function of automatically creating a story structure by the user only by setting some requirements by the user, thereby realizing the conventional computer software. It is a way to further improve the way of the wear, and based on the above-mentioned zero technical background, it goes beyond the mere extension and improvement of the existing technology, and this book! It can be said that it has unique novelty.
  • a basic method that can integrate the knowledge necessary for creating a story structure into the process of creating a story as a tree structure (hereinafter referred to as a “story tree”) is used.
  • the creation of the story structure corresponds to the expansion or transformation of the story tree, and the mechanisms for these operations are central to the methodology.
  • Such a structure is configured based on the structure. Of this, shouting that of the former hand * 3 ⁇ 4-out 3 ⁇ 4> knowledge and "The Story technique," said Kohi, ⁇ product ⁇ that of the latter-controlled knowledge ".
  • the narrative branch method requires a reference to a knowledge base that stores «minds and other knowledge» for the narrative expression.
  • the ability to use a fragmentary knowledge group to enable the flexible automatic generation of stories without having to trace the first-speed steps to create a specific story structure in advance. can get.
  • the creation of a story structure corresponding to the connection of the terminal nodes of the story tree in the above story and plot means such an extension or transformation of the story story.
  • the story-making section the main part of the story is expanded, and in the professional / 's story-making section, the story tree is mainly transformed into a story. Narrative technique means the means to concretely execute these works.
  • each narrative technique which performs an extension or transformation of narrative II in various styles, defines how the narrative tree is manipulated in a particular way, each of which is a fragmentary piece that performs a single task. It constitutes a simple handicraft module.
  • a networked knowledge base is consulted.
  • the expansion of the story mainly creates another event or subtree that has some M factor for a specific event or subtree that constitutes the last part of the story, thereby extending the story tree as a whole. It is done through doing. Recognition of such an entity for a certain event or subtree is performed through a network-like knowledge base.
  • this system divides the knowledge used in creating the narrative structure into three main parts: the narrative technique, the narrative update, and the knowledge base. Utilization is handled by the rules of the story strategy, which means the control mechanism of the entire story structure creation process, without having to write a fixed algorithm that matches the creation of a particular type of story in advance.
  • the rules of the story strategy means the control mechanism of the entire story structure creation process, without having to write a fixed algorithm that matches the creation of a particular type of story in advance.
  • the conventional story generation system is often positioned as an experimental system for verifying a specific ethical framework in the information processing field such as artificial intelligence. It is based on the principle of operation limited to the theoretical framework of the above, and aims to be a basic technology for constructing a new framework of computer systems by analyzing the structural characteristics of the story itself from many aspects. As a result, nothing was configured as an integrated system. In addition, these conventional systems aim at verifying the rationale in artificial intelligence through a simple simulation of human behavior in a very small story world. (The real human world is too complicated. This is a virtual realization of the world of object 1 ⁇ for the experiment), and did not bring a new architecture on the system.
  • the diversity of the story development is only a mere selection from several possibilities fixed by the creator in advance. Do not have the ability to create stories at all.
  • the present invention realizes a completely automated mechanism for creating a story structure by introducing a story strategy based on story parameters, and in that sense, adopts a conventional approach such as simulation and simulation. On the other hand, it is an inherently different approach.
  • the S-type connection between the “marketing advertisement automation support system”, which is a type of invention previously submitted by the present inventors, and the present invention is as follows.
  • the system creates a series of event sequences (scenes) in which a specific segment (person type) acts, and uses it to create a complex
  • scenes event sequences
  • person type person type
  • a fixed algorithm consisting of a combination of rules is described in advance as a normal procedural processing procedure, and the above rules are limited to a small number and have a mechanism depending on the specific field of advertisement .
  • the types and mechanisms of story techniques for creating a story structure are greatly enhanced and expanded, and a unified principle of expansion and deformation processing at the end of the story is established.
  • the types of material and techniques are classified and systematically organized.
  • some of the rules overlap with the rules in the above-mentioned marketing advertisement automation support system.
  • the application of the technique has been strengthened and expanded compared to the conventionally proposed rules W, and many other branch methods have been added and expanded.
  • Kibatsu BB the most important novelty in Kibatsu BB is the creation of a story structure.
  • the thing that controls the entire process. By elaborating the collection of rules called iSdft abbreviations. It is possible to create various types of narrative structures without having to pre-write the target argo-sum. This is a mechanism that does not exist in the conventional marketing advertisement automation support system based on the fixed draft algorithm. In other words, the creation of the story takes into account the requirements of the user at the time. The rules of the story war, or the rules of the simple story branch and the end of the story can be redeemed. The e- equipment-and-branch method group executed by will function more effectively only in conjunction with this automated mechanism.
  • each moshur is composed of a set of fragmentary procedures and rules, it is easy to add and delete these rules, and they do not make the system inoperable. It also has the property of This is associated with the advantages of easy scalability and changeability of the system. These characteristics are also the characteristics that flexible automation control based on the above-mentioned story moth scheme is achieved by controlling the entire process.
  • the present invention is based on the existence of a flexible control mechanism, that is, a story strategy, at the core, and speeds up compared to the method in the conventional marketers' advertisement automation support system. It has features such as expansion and systematicity, system expandability and changeability, versatility, etc., so that new functions can be realized.
  • a flexible control mechanism that is, a story strategy, at the core. It has features such as expansion and systematicity, system expandability and changeability, versatility, etc., so that new functions can be realized.
  • the system automatically creates texts with a story's structure to create various types of users.
  • a system that can be used for creative work and can be used as an advanced human interface in various computer systems. Did not exist.
  • a scene in which a user conceives an outline of an easy-to-use or business story and an idea line can be assumed.
  • the system is required to have the ability to create stories of various configurations in winter, but at the same time, the ability to narrow down to a small number of stories and present it to users according to the needs of users. It is necessary to have the ability to easily test different possibilities.
  • a general purpose system that can be easily expanded or changed according to the type of story desired by the user and the purpose of use is also required. (Note that the usage status is also used in a system that requires a kind of simulation, such as education or games.) Or calculate the progress of the simulation automatically while using the system.)
  • Narrative structure creation * The present invention also envisages the possibility of generating a story or acting as a kind of stimulant for human creative or imaginary tasks. There is a need for a method that can easily and flexibly realize a new story. In order to solve such a technical section S, it is impossible with an algorithm limited to the creation of a story of a specific structure in advance, and many combinations can be easily realized from a set of more fragmentary elements. We need a mechanism that can create
  • an event is a description of an event that is composed of ginseng (moving subject, object, time, embankment).
  • the material information is (Kill Ichiro Kozuki 1 midnight 1 place 1). This is "Ichiro kills Kozuki at location 1 at midnight 1”.
  • each element in the list is a specific narrative in each knowledge base.
  • the system corresponds to an item, and a specific data value is stored in the item.
  • the user sets values for a plurality of parameters (referred to as ⁇ story parameters overnight j) as a description of the required items that are used in the strategy for creating the story structure.
  • the system will automatically create a name from then on.
  • the process is divided into three stages: core story one story story one story plot.
  • the processing of these three-step swords is all realized by the same excision mechanism, but different physical structures are created depending on the amount of knowledge units used and the order of use. , Its construction is mainly caused by the achievement of the above parameters.
  • the system is roughly divided into four parts: the object creation management module, the object technique module, the object path module, and the knowledge base module.
  • the basic way to create a story is The rules stored in the word strategy module determine the specific material techniques in the narrative technique module to be used by referring to the values of the narrative parameters that were initially input. This is done by repeating the cycle of executing a specific part of the story and expanding the story.
  • the story moth road is best suited for each of the following stages: core story creation, story creation ⁇ pro, and story creation, and what is the value of the story parameter. Select a narrative technique and implement a specific narrative technique for any part of the story that is being developed.
  • a core story is the first template of a story that results from applying a single narrative technique to material information.
  • a narrative technique that develops the minimum basic pattern of the story corresponding to the theme of "murder by heritage" is used.
  • using different narrative techniques creates another core story from the same human power.
  • Hayakawa owned a huge amount of property.
  • Hayakawa had Ichiro as a child.
  • Hayakawa had a child, Kozuki.
  • Ichiro wanted to monopolize Hayakawa's property.
  • Ichiro killed Kozushi. Ichiro robbed Koizuna of money, Saya exposed who killed Kozuka, police searched for -ro, police confirmed Ichiro, and c.
  • Storiichi means an era based on the chronological order of events.
  • time progresses straight from the first event to the ft ⁇ ⁇ event.
  • the story is created by repeatedly applying several narrative techniques to each event in the core story, and is an art line that follows the time sculpture of the entire story.
  • -6- It is structured as a successful and poor narrative structure based on a particular theme.
  • a story structure is created from a single event group based on the differences in story parameters in the order of core story, story, pro, and story.
  • the knowledge processing required to create a story structure is integrated into the process of creating a story structure as a tree structure (hereinafter referred to as a “story tree”). It is constructed in an integrated manner, focusing on techniques that can be embedded.
  • the creation of the story structure corresponds to the expansion or transformation of the story tree, and the mechanisms for these operations are at the center of the method wheel.
  • Such a mechanism is constructed based on two types of knowledge structures: manual knowledge for expanding or transforming the story tree in a convenient way, and rule-based knowledge for controlling its use. .
  • Handicraft requires a knowledge base that stores concepts and other knowledge for the narrative expression.
  • the process of creating a story structure can be roughly divided into two parts: the creation of the conceptual structure (herein called the story structure) and the creation of the surface s-words.
  • the story structure means the internal representation of the intentional structure in the story within the computer.
  • the invention treats the term narrative as follows: e. Narrative simply means not only the narrative as a surface natural language level, but in a broader sense, including its structural aspects.
  • the story in the meaning is used in correspondence with the English word N a. Rr ⁇ tive ⁇ ⁇ ⁇ And the whole story in such a broad sense is divided into several structural levels and considered.
  • Each story , Professional, '/ Bok, co-Nsu Torakushiyon are to be referred to by use of special ⁇ that Tex me.).
  • Figure 1 shows the basic method of creating a tank structure. The technical points of this method can be summarized in the following five items.
  • a story is a structure based on the temporal pyogenesis of an event
  • a plot means a structure that is transformed according to the structure of an expression as a story.
  • the process of creating a story structure as a physical tree and its work The conceptual structure of a story is expressed as a story tree with one vertex at all levels.
  • the terminal nodes of the story tree correspond to individual events.
  • the other nodes correspond to the relationships that control the lower nodes.
  • stories and stories are a quick link between the end nodes of this story.
  • Creating a story structure is an extension or transformation of such a thing. The main part is to expand the story tree, and in the part of creating the project, the c- story technique, which mainly performs the transformation of the developed narrative as a story, It means a specific action to be taken.
  • Each narrative branch method defines a specific way of manipulating the end of a narrative, each of which constitutes a fragmentary handicraft module responsible for a single task.
  • the main narrative technique for expanding narrative II is to use various types of relational concepts that can serve as nodes in the narrative tree, and to use events and partial segments in the narrative tree (hereinafter referred to as “partial end” in the narrative tree). Is called).
  • the narrative technique for transforming botanical trees is based on a story tree that has been composed once. Using a viewpoint of a specific appearance or a pattern of development of a specific genre, etc. Add operations.
  • a story moth scheme that controls the use of narrative techniques narrative knowledge that means control to determine at what point in the story tree to apply which narrative technique to which part of the narrative tree Call.
  • Narrative strategies are defined in fragmentary rules and are driven by control mechanisms.
  • the description of the main rules refers to the story parameter j and the current state of the story tree that specify the characteristics of the story to be created in the condition section, and specifies the story technique to be executed in the conclusion section.
  • the system will automatically create a story based on those requirements.
  • the story structure of the form desired by the user is flexibly created.
  • Creating a story structure using a network-like knowledge base The expansion of a story is mainly based on a specific event that dazzles a part of a story tree or another event or subtree that has some relation to the subtree. This is done through the creation of a story tree, thereby extending the story tree as a whole. Recognition of such relationships for certain events or capital-ary tree, the value is performed through 7 Bok work-like knowledge base.
  • each element (event, thing, thing, etc.) that constitutes the knowledge base has multiple characteristics, and this component and the description of another component that has a specific M
  • the narrative technique creates events and subtrees that have some kind of M-factor with the events that were the target of the work, through the imaginary W-factor of the nature of the character.
  • Create Core Story Create ft's first story tree from a set of one or more fragmentary input events.
  • a fragmentary event set means a set of events that do not constitute the end of the story.
  • the story tree first constructed in the story creation described below is especially called the core story in the sense that it is the first unit of ft in the creation of the contradictory structure. Same as creation.
  • Pro creator: Transforms a story tree that follows the temporal progress of human interaction into a sculpture that follows the composition of the sculpture, or text, of the narrative. Therefore, it corresponds to the creation of the expressive aspect of the story with respect to the creation of the story content.
  • task B leaves instead of g leaves.
  • the story to be processed is made into a story, and the processing of that part is handed over to the next project / creation, and when that is done, the task returns to the story creation task again. It is assumed that a pass is also possible. Or, first, conceive the thing ⁇ expression, that is, the professional ", u, and * f It is possible to create a good story content, story 1 / one.
  • the conceptual structure of the story is expressed as a story tree at all levels.
  • the terminal nodes of the story tree correspond to events, and the other nodes control lower nodes by some M factor. This is called the “r-relationship”.
  • the sequence of events from left to right at the end of the story is equivalent to the core story, story, pro, and book, respectively.
  • the target tree takes the form of a hierarchical list.
  • each event is expressed as a list containing several pairs of sexes and sex values, with the verb concept (called the r-event concept j) first as follows: (For the notation, the following example shows one example. If the above conditions are satisfied, other methods may be used.) ⁇ Characters include characters, time, places, objects, etc. e The factors that can be the IS nature of these events are called r-story-off'sect.
  • Entities are high-order nodes for one or more low-order M or low-level events, as follows. This is shown in FIG.
  • One event can be made into a different story tree using a few different relationships. For example, (Eat ((subject Taro) (target breakfast) (time morning) (place home)) ("Taro eats breakfast at home in the morning. J) A different tree is formed by combining with the events and sub-ends of the M that make up the M-factor.
  • the attribute of the event concept and other attribute values that make up the event have corresponding items in the knowledge base, and specific values are described in them.
  • the name of each item of the character knowledge pace, the object knowledge arrow base, the place knowledge base, and the time knowledge base is described as the gender value of the subject, subject, place, time, etc. that constitute the event. .
  • attribute values such as date of birth and gender are described in the item “Taro” in the character knowledge base.
  • Story making is done through such expansion and transformation of the story tree.
  • a skirting tree is expanded by combining a particular event or branch tree in the story tree with another event or tree that has some M factor.
  • the transformation of a physical tree is performed by reorganizing the narrative tree or subtree from a specific viewpoint.
  • Story expansion is basically possible for any node in the story tree.
  • An extension of a story tree to a node other than the terminal node means an extension to a subtree of the story below that node.
  • story transformation is basically possible for terminal nodes, that is, for any node other than a single event (ie, a subtree of a story).
  • story technique j the intuitive knowledge for expanding or transforming a story tree is called "story technique j.
  • a story technique is basically a single task.
  • the narrative branch method is a type that extends a narrative tree, it is defined as one physical technique for each type of relationship that is valid for a certain event.
  • the narrative techniques include not only the type that creates such a narrative tree, but also the elements that change the character of the narrative object, that is, the components of characters and objects used in the end of the narrative. In the end, any kind of intellectual knowledge used in the creation of a narrative structure, centered on narrative tree operations. Each of them is basically a narrative technique that performs a single task.
  • the story structure as a whole On the other hand, the process of making a story shall be performed by a combination of the use of narrative techniques, but in the practice of using narrative techniques, it is natural that the strategy differs depending on the type of story to be created.
  • Different kinds of knowledge are required to control the use of the narrative branch method, for example, the question B, which narrative technique is used for which nodes in the disjunction tree, is solved in the actual creation. need to come out.
  • ⁇ story iS shown the knowledge to make a decision process in the control for the story branch method ".
  • the storyline iS must be used.
  • the basic policy of what kind of object you want to create needs to be determined, and it is expressed as r object parameter j. .
  • the story SS stands for knowledge to control the specific use of the story branch method according to a specific story parameter combination.
  • Ss »Abbreviation is a ⁇ d report that shows the story of each story and the current status of the end of the wis. It is expressed in the form of rules with the conclusion of the story technique to be specified, and has a separate controller for executing rules and resolving conflicts.
  • the method of creating a story structure is thus composed mainly of two concepts, story technique and story strategy, but a knowledge base is also an essential component.
  • a particular narrative technique creates another event that has some relation to one event in the object-reception tree, refer to the sex information of the components, including the concept of events in the knowledge base.
  • the knowledge base exists mainly as a means for narrative techniques to provide basic data for expanding and transforming story trees.
  • each narrative technique is not a technique that defines knowledge for a specific type of story, but a fragmentary technique that defines general narrative structure creation knowledge, albeit with different degrees of abstraction. It is a flexible method with versatility and extensibility, as it constitutes a framework that can be added and defined and added in combination. .
  • defining a narrative schema specific to a particular type of story This is possible not only through narrative techniques but also through narrative moth schemes. For example, in a case where a story with a shorter and simpler structure or a story with a remarkably impressive structure is required, setting values for the length and complexity of the story parameters and defining rules that reflect them are necessary. You can knot.
  • the method of wood invention solves the technical problems mentioned above *.
  • tactical knowledge corresponds to knowledge for the task of creating a deep conceptual structure
  • branching or tactical knowledge corresponds to knowledge for the task of creating a superficial natural B word.
  • each level of the story tree, ie, story creation, and story expression, ie, “pro”, and story creation in the creation of the IS structure is actually manipulated in some way to manipulate the story tree.
  • a hand-knotted weir to be a forensic knowledge
  • the argument of the object technique for expansion is an event or a subtree.
  • a new event or tree having this argument and some IWI relation is created, and a tree in which the user is connected by this M relation is returned.
  • the transformation-object branch method uses the following: object: *: The whole or part of the argument is returned as an argument.
  • Other types are those that combine or tie together multiple arguments, and those that change the attribute value of a particular narrative oashect contained in a tutor. Keyword is used, for example, if you want to specify a special case divided into ⁇
  • the object 1 ⁇ branch method is applied to a specific node of the story tree as an argument, and returns a new value obtained by expanding or changing the node. And update the end of the story as a whole by replacing it with this new value.
  • the application of the narrative technique is not necessarily determined at the terminal nodes of the narrative story, but may be at other nodes, that is, subtrees. There is also a feature that it becomes possible. While the terminal nodes in the story tree are progressively refined, it is possible to create them manually, but on the other hand, the story to the node of the story tree '.' If you apply the technique repeatedly, the story tree will gradually grow upward. If the former corresponds to the story of the story, .'butown generation (Fig. 7>), the viewer corresponds to a kind of bottom-up (Fig. 8>). The choice of nodes to use a certain e- story technique is done through the execution of a story moth scheme.
  • Narrative technique using a local cabinet between concepts Create another event or a branch tree with a specific local relationship with the event or part end of the argument, and return a value that combines the two.
  • a narrative branching method that uses knowledge that is used as a local relation between concepts such as “reason” or “cause” relation, and “result” relation.
  • the definition of this kind of local M-factor it is mainly described in the context of the event concept in the arrowhead base.
  • Narrative techniques for creating narrative objects x-cuts characters, objects, places. Time-of-day narratives that appear during events, such as changing the attribute values of r. Or create a new instance of a story object. For example: A. Change the age of a person or set the character's current occupation. This is done for the corresponding climb in the knowledge base linked by the link.
  • Narrative techniques using the method of the narrative of the narrative Extensions aimed at impressing the narrative-There are physical techniques that use the art of the initiative for transformation. For example, reconstructing a story or a part of it from the viewpoint of a specific character who climbs there; There are narrative techniques.
  • Narrative techniques as the definition of a narrative schema For some types of narratives, certain narrative techniques can be applied at a fixed location. In such a case, a plurality of narrative branch methods can be combined into one independent object S technique, which is a framework that defines the method of narrative. Call.
  • the article advisory strategy includes the next node in the current story tree to be expanded or transformed (this is called the “story branch method application point” or “growth point”), and There are two types: those that determine the narrative technique to be applied to the node.
  • the narrative PS consists of a set of these two types of rules and a control mechanism for controlling these rule groups.
  • Each rule in the story strategy consists of a conditional city and an execution unit.
  • the condition capital is mainly composed of parameters that define the betting characteristics of the story to be created, and these parameters are called r story parameters.
  • the execution unit there is a K predicate of a node that can be used in a feasible proof technique or a word branch method, that is, a point for a story technique.
  • the story parameters used in the condition part of the rule include * theme *, * length *, * character uniqueness *, * »article *, * view *, * poma * other species There is an amount, and the value is set by the user at the beginning of the story creation.
  • Each story parameter has a plurality of value candidates.
  • * length * is long / medium / short
  • * character unique figure * is ⁇ Z medium / low.
  • Accordance connection as such sets created story in such a way as to ⁇ the value of the story parameters is performed, there are necessary things 3 ⁇ 4 ⁇ rule is configured e
  • the first type of narrative strategy the method of determining the point of application of narrative techniques, can be any of the following concrete implementation methods, from simple balance to more complex control.
  • Simple formal method This is a method of determining the application point of the story technique by changing the nodes in the story tree in a formal manner at a certain level. For example, move the narrative branch method point from the terminal node (event) at the left end of the story tree to the right with I-spurt, and when it reaches the right end, repeat the same processing for the newly created event ( Figure 9) (Similar processing may be performed from the right pheasant tree to the left.) It is possible to use the method as a point of use (Fig. 10) or a combination of the two methods. There is a method of intensively processing only the terminal node of.
  • Focusing method It is a method to set the focus to be expanded intensively in the skirting tree. For example, it is conceivable to set a particular character as the focus, and to focus on the event in which it appears as the application point of the physical IS technique. By switching the focus during the creation, you can create a story part centered on another character or other story object.
  • More complex strategic method refers to multiple story parameters, the current story tree shape, etc., and determines the optimal story branch method point for each point in the creation.
  • the above two methods can also be regarded as two alternatives in such an IftBS-like determination method.
  • the narrative to drive the narrative technique is based on the value of the narrative parameter, the amount of the current task such as story creation, pro, and zol, the type of the target node (event or part).
  • Information such as the node that controls the branch tree) or the position of the target node in the story tree (mainly the depth repertoire) is required.
  • the node that controls the branch tree or the position of the target node in the story tree (mainly the depth repertoire) is required.
  • the * theme * is a story, it is thought that the whole story needs to be supported by this * ema *.
  • the subsequent narrative composition would have to be done exclusively for nodes other than the top-level nodes. Because this type of information may be needed, the conclusion of the narrative strategy rule should specify a specific narrative technique and be able to describe information that will serve as constraint knowledge for subsequent creation. I will need it.
  • the physical parameters are classified into those related to physical structure, those that apply M to the world structure (characters, space, and time), and those that apply M to expression.
  • the parameters that are categorized into the story structure are directly related to the composition of the box tree itself! Echoes.
  • WI which is the main part of the story.
  • the * theme * mentioned above implies the use of a particular theme pattern
  • the * narrative * controls the use of knowledge to develop events specific to the IS, such as entities.
  • the expression strategy is a parameter that is more strongly related to the narrative of the narrative expression.
  • * viewpoint * is a specification of the type of viewpoint that is the main subject of expression of the object, and this affects the structure of the professional ".
  • the world structure specifies the types of narrative objects and other characteristics of the narrative composed of events. However, if W example there are also parameters on the saturation »the opportunity 3 ⁇ 4 at the end of the story * registered person ⁇ -. 3R 0 ⁇ * eyes in W ⁇ .” - plan M engagement and M Hayashi and are r 3.2.7.2 Story structure creation cycle
  • the creation of a story structure based on story strategies and techniques is basically accomplished by repetition of cycles as shown in Fig. 11.
  • the first type of narrative, 3 ⁇ 4 then applies some narrative techniques to determine the nodes to be expanded and deformed, that is, the application points of the narrative techniques, and then the second type of narrative strategies Determine and execute the story branch method to be applied to this. Then, the execution result is combined with the story tree as a whole, and the cycle ends. This is repeated until the termination condition is dropped.
  • this process is performed from the creation of the core story to No. I. (In this case, we decided on the point of the material technique and the technique, and then decided the narrative technique to be used quickly.
  • next narrative technique or candidate to be executed is determined first, and then the applicable point of the narrative technique is determined.
  • selection of the story is also determined by the story strategy.
  • the processing of the story strategy includes resolving conflicts between executable rules, processing rules that have been used once (reusable, unusable or priority processing, etc.), and end conditions for each task of creation.
  • Process of creating IS structure The process of creating a story structure is defined as consisting of three tasks: core story, story creation, and plot creation. Of these, the former two are responsible for the creation of the story contents, while the professional and zoto are responsible for the creation of physical expressions. All of these are treated as a unitary process of expanding or transforming the story tree.
  • Narrative expression-Narrative tree A narrative is expressed as a narrative tree in which the event is the final contraction node and the subordinate nodes are the other nodes. Each event consists of a list of attribute-values of event concepts and other narrative object concepts, and the event concepts and attribute values are linked to the knowledge base components *.
  • Physical structure creation method Physical structure creation is performed through expansion and transformation of the story tree.
  • the expansion of the object tree corresponds to the creation of the content of the story, while the transformation of the story tree corresponds to the creation of the object box representation.
  • These processes are performed by a narrative technique, which is a set of fragmentary handicrafts defined for performing operations on the object a tree. In doing so, an object consisting of fragmentary rules and their control mechanisms controls the use of narrative techniques.
  • the narrative structure is created by repeating a cycle in which the narrative warfare determines the point of the narrative technique in the bargaining tree, executes a specific narrative technique, and combines the result into the narrative tree as a whole. Done.
  • the method of the object-branch method The function of the narrative technique is as described above, but it is a technique that uses the local relationship between concepts.
  • the technique of creating a narrative object, the rhetorical branch of a narrative It can be divided into five types: techniques using techniques, techniques using macro structures of «» » and techniques as narrative schemas.
  • the article advisory law can be freely defined and added.
  • This output gives the input fragmentary phenomena meanings such as causality, refinement, and context, and also provides the user's request in the form of story parameters.
  • the story structure of the tie corresponding to them is automatically created. Users can use this output story machine as a basic data. It can be further converted to a detailed surface expression such as natural language, or used as a hint for creating various stories. It can be used to support ideas. It is also possible to use this system as a narrative human interface for educational systems. By simply changing the settings of the material parameters, different types of narrative structures are automatically created, so that it is possible to repeat the work * for each of these various purposes in a short period of time. It contributes to improving the efficiency of use, supporting creativity and imagination abilities, and making the interface friendly. BRIEF DESCRIPTION OF THE FIGURES
  • Figure 1 is a diagram showing the outline of the method of the story structure creation automation support system.
  • Fig. 2 is a diagram showing the process of creating a story structure in the automated story structure creation support system.
  • Figure 3 is P1 showing an example of a story tree format.
  • FIG. 4 is a diagram showing another example of a story tree format.
  • Figure 5 is a diagram showing an extended form of a story tree.
  • Figure 6 shows how story trees of different formats are created from the same input.
  • FIG. 7 is a diagram showing a state in which the story tree is expanded in the lower direction.
  • FIG. 8 is a diagram showing a state in which the story tree is expanded in the upper direction.
  • FIG. 9 is a diagram showing, as an example of a technique for determining the application point of the story branch method, a method in which only the terminal node is used as the application point first.
  • FIG. 10 is a diagram showing another example of a technique for determining a point of application of a story technique, in which a top-level node is used as an application point in a track number.
  • Fig. 11 is a diagram showing an outline of a story structure creation cycle in the object structure creation automation support system.
  • Figure 12 is a diagram related to the system configuration and the processing flow, showing the system configuration and input / output relationships in the port structure creation automation support system.
  • Figure 13 shows the operation style of narrative thick by the combined narrative branch method for one event.
  • Figure 14 shows an inclusive narrative branching method for events in a basin! Is the operation format at the end.
  • Figure 15 is a diagram showing a narrative operation format using the element-substitution type-branch method for a single event.
  • Figure 16 is a diagram showing the operational form of an object
  • 17 is a diagram showing a narrative type of narrative operation based on an introductory narrative branching method for phenomena.
  • Figure 18 is a diagram showing the operation style of story tree by (1) deformation for one event, (2) story technique.
  • FIG. Fig. 21 is a diagram showing the operation format of a story tree by an element replacement type story technique targeting a subtree.
  • Fig. 22 is a diagram showing the operation format of a story tree by the argument substitution type story branch method for a city tree.
  • Figure 23 is a processing flow diagram showing the procedure for generating a story by the automated story structure creation support system.
  • Figure 24 is a diagram in which the opening scene in the novel of the analysis example is composed of a story tree.
  • 0 25 is the first figure in which the novels of the analysis examples are constructed as professionals and / or trees.
  • Fig. 26 is a second diagram in which the novel of the analysis example is similarly configured as a professional's tree.
  • Fig. 27 is the third figure, which is also composed of the novels of the analysis example as professional and ⁇ trees.
  • Fig. 28 is the fourth diagram in which the novel of the analysis example is similarly constructed as a plot.
  • Fig. 29 is the fifth diagram in which the novel of the analysis example is similarly constructed as a professional and a sot tree.
  • Fig. 30 is the sixth diagram, which is also composed of the novel of the analysis example as the end of the program.
  • Figure 31 is the seventh diagram, which is also composed of the novel of the analysis example as a professional, ',' tree.
  • Fig. 32 is the eighth diagram, which is also composed of the novel of the analysis example as a professional.
  • Figure 33 is the ninth diagram in which the novel of the analysis example is similarly constructed as a plot tree.
  • Figure 34 is the 10th figure, which is also composed of the novel of the analysis example as a professional tree.
  • Fig. 35 is the first diagram in which the novel of the analysis example is similarly constructed as a plot tree.
  • Figure 36 is the first figure in which the novel in the analysis example was configured as story thick.
  • Figure 37 is the second diagram in which the novel in the analysis example is configured as a story tree.
  • Figure 38 is the third diagram in which the novel of the analysis example is configured as a story tree.
  • Figure 39 is the fourth diagram in which the novel in the analysis example is configured as a story tree.
  • Figure 40 is the fifth diagram in which the novel of the analysis example is configured as story II.
  • Figure 41 is a diagram showing the deformation relationship between the plot tree and the story tree.
  • Fig. 42 is a diagram describing, in a physical tree format, an example of a core story, that is, an example of a core story 1> created by the material structure creating execution.
  • Fig. 43 is a horizontal view showing the process of creating a physical structure by using the real K.
  • Table 1 shows the main species 18 of the story branch method and their functions! ! This is the first table arranged as follows.
  • Table 2 is a second table that lists the main naming methods « «! And their respective functions.
  • Table 3 is the third table that also sorts the seed value of the main ⁇ -branch method and each function as one R.
  • Table 4 is the first ft table that lists the story parameters »and the main items
  • Table 5 is a second table that lists the story parameters, the types of the main story parameters, and their definitions.
  • Table 6 is the first ft table to show the analysis results (summary at the plot level) of the plots and plot structures of novel works taken as analysis examples.
  • Table 7 is the second table illustrating the analysis results of plot structures (summary at the level of the box 'and the box) in a novel work also taken as an analysis example.
  • Table 8 is the third table that illustrates the analysis results of the professional structure of the novel work, which was also taken as an example of analysis.
  • Table 9 is the first table that illustrates the analysis results (summary at the story level) of the story-by-story structure of a novel work taken as an example of analysis.
  • Table 10 is the second table illustrating the results of the story-by-story structure analysis ("summary at the story-level") of a novel that was also taken as an example of analysis.
  • Table 11 is the third table illustrating the analysis results (summary at the story level) of the story structure in the novel work also taken as an analysis example.
  • Table 12 is the fourth table that illustrates the results of story-structure analysis (summary at the story level) in a novel that was also taken as an example of analysis.
  • the story technique expands or transforms the story tree in whole or in whole under the control of the story ift, against the background of a network-like knowledge base, and the story creation management module summarizes the result as a whole.
  • Create a narrative structure by repeating the cycle of integrating with the thing advisory tree.
  • the automated system for creating a story structure is composed of the following four main modules.
  • Story creation management module First, set the initial values required for the core story, story, and plot creation tasks. Then, a story tree or a fragmentary event group that is currently being processed is given, and the processing is delegated to the name moth abbreviation module and the story technique module. When a merchant is returned by the rain module, it usually expands or transforms the object IS tree by replacing this ancestor with the nodes in the current narrative tree that have become points for the narrative technique. This process is repeated until the end condition of each task is satisfied.
  • Physical strategy module Includes physical strategy rules and their control mechanism (physical strategy control mechanism). Once the point of narrative technique success and the narrative branch method to be executed are determined, the execution of the object branching method is entrusted to this narrative technique in the narrative technique module. This module realizes automation of story structure creation.
  • Story branch method Contains a library of story techniques. At the point where the narrative technique is determined by the Compensation Law strategic rules, the material techniques in this library also determined by that are used.
  • Knowledge base module Seven types of concept arrowhead base of each concept of event, thing, thing, condition, person, place, and time, script, causal M, goal total, ⁇ , mental It consists of a rule-based knowledge base for reactions and theme structures.
  • the latter type of knowledge base is referred to herein as the "story knowledge base.”
  • the concept knowledge base is a set of frame-form concepts, and the link between the concepts is a network. It is stretched.
  • the narrative knowledge arrow pace is a set of rules for creating a concept with a particular concept and another concept with a specific relationship with it. This rule name is mainly described as the slot value of the event concept in the concept knowledge pace.
  • the experimental system is described in the programming language IfCommonLisp.
  • C ommon L isp language Ri beneficial languages der for list processing, such as. Strings, but c are used often in artificial intelligence research, the present algorithm is rather a language-dependent, the other programming Nge Language It is of course possible to realize it.
  • each component in the system is agreed as follows.
  • the name of each branching method is expressed as tech ⁇ proper name>.
  • the notation for the name of each story strategy rule is str & ⁇ unique name>.
  • a frame in the conceptual knowledge base (the name of an abstract type, such as "station") is, for example, objS! Beer.
  • the symbol that represents the type of concept (here, obj) and the unique name are ⁇ symbols.
  • instances names of concrete values, such as “Kyuko DT” for the above “ku” use the two connected with an & symbol, such as ob.j & beer1 .
  • Human power into the system is (1) a set of values for the story parameters and (2) a set of one or more fragmentary events, ie, events that do not form a tree structure. These are entered first by the user. (2) means the material that first triggers the creation of the system's narrative, rather than the semantic content that should be spoken by the system, and is referred to here as "" material information "”. Based on these, more and more core stories are created, and then stories and pros are created. In the current implementation, these are processed step by step with this tongue. However, looking at each task unit does not necessarily mean that the story is elaborated from the top down. As noted in the previous chapter, when expanding or transforming a story as a story tree, it has the characteristic that any node in the story tree can be processed equally.
  • the heels of story creation in each task are as follows. First, when the user manually inputs the material information and the value of the story parameter through the plow of the story creation management module, the control is transferred to the material ⁇ & brief controller «and the core task, which is the first task of ft, is created. .
  • Outline The rules of the story branch method to be applied to the material information are determined by the rules, and the e -coast story creation task that executes this task. Processing ends once the story end is constructed. . In the task of story creation and brochure creation, the variables required by the story creation management module are used. After the value is set, control is transferred to the story strategy control mechanism.
  • the story technique in the story technique library is executed, and the result value is again processed by the story creation management module in the story tree as a whole.
  • the story tree as a whole is expanded or transformed.
  • the cycle of the narrative determining the application point and the narrative technique by the narrative is repeated until the end condition is satisfied. In this way, in each task, the processing of determining the application point of the story branch method, determining the story technique, executing the story technique, and expanding and transforming the entire story tree are performed in common.
  • the knowledge base is broadly divided into a concept knowledge base and a narrative knowledge base.
  • the concept knowledge base is used to store the attributes and values of various concepts used as components of events in a story.Each concept is a set of lists consisting of sex / value pairs. Defined. There are two types of concepts, those that mean the generic name of a class and those that mean its concrete substance. Here, the former is called a frame and the latter is called an instance.
  • Conceptual knowledge bases are categorized into the following types.
  • Event concept knowledge base A knowledge base that stores verb concepts. The concept of events is an essential element in the composition of all events, and is the most basic purple essential for the development of a story.
  • Knowledge Conceptual Knowledge Base This is a knowledge base for the concept of a physically existing thing, that is, a substance. In many cases, the concept of a thing and the concept of a thing represent two aspects of the same concept (for example, a school as a building and a school as a matter). Include in knowledge base. Equivalent to a noun with physical mass.
  • Thing concept knowledge base A knowledge base for abstract things (things) that do not physically exist. It is represented by the noun form of the verb concept.
  • Modified Concept Knowledge Base A knowledge base of concepts for modifying other concepts, and is equivalent to adjectives and adjective verbs.
  • Person concept knowledge base Knowledge base that stores instance information of characters. Each instance sweeps into an occupational frame on a knowledge-based basis.
  • Time concept knowledge base A knowledge base that stores time instance information. The information that is used for time divisions such as morning and noon is included in the concept knowledge base.
  • Space concept knowledge pace This is a knowledge base that stores instance information on spaces such as specific lands and shops.
  • R be " ⁇ column of the definition of the frame of the gutter ⁇ event concept.
  • evc kill is the name of this frame, where "kill” means the form displayed during the event ⁇ :.
  • attribute-value pairs are described.
  • each of the properties from agt to tim can be used as the required purple when an event is created.
  • an event force such as (kill (agt agt & ota-nobimaga) (ob.j agtisuzuki-yusule) (timt & asal) (loc loc & suzuk u ke)) is composed.
  • Attribute values are frames in the concept knowledge pace
  • the previous example is an instance of the instance of the person to be killed, named agt & siizuki-yusuke. This is equivalent to the instance of the frame named thX child included in the concept knowledge base.
  • thX child included in the concept knowledge base.
  • Each) »The value described for gender may be the name of another component in the knowledge base (for example, thX elementary school) or a simple constant (for example, Suzuki).
  • the definition of an event concept that is causally related to a certain event concept it is necessary to create different event concepts depending on the difference in attribute values in the vehicle phenomena including the event concept, and this must be created in the event concept knowledge base. If it is directly expressed as the attribute value in, it becomes popular. Therefore, the base of the definition of such a complex relational concept, the attribute value of a specific concept describes the name of a specific item in the word knowledge base, and by referring to it, Let's create the concept of M-person. For example, in the example above, the causal M (result and reason clerk).
  • Script Knowledge Base Includes rules for creating scripts that develop target events as sales of lower-level abstract events. The script, if t e is supervised by a succession w engagement, "" strangle to death "that event, pouncing r, strangle, suffer, choking to. To ⁇ , are deployed on the speed chain that is the script of the event of death”.
  • Theme pattern knowledge base Contains the rules for creating main patterns that include certain events. For example: "One version of the theme, Murder by Succession J, owns a ⁇ parent or ⁇ large property, and one of the children kills his brother to monopolize it or the murder is exposed. He is arrested. "
  • Story-related knowledge base Contains rules that create a bank of events based on a story and developed from the story.
  • the story W engagement "prohibition - violation", ⁇ perpetrators - Fixed “,” instruction - compliance ",” Problem One solution ", ⁇ desire - sufficiency”, and the like.
  • the above-mentioned “r-building-plan” relation can be considered as a kind of story M section.
  • ⁇ In one example of the offense-elimination j relationship,
  • a pattern is created that consists of “recovering” (dissolving) the (perpetrated) object and awarding the reward.
  • Compound story-related knowledge base Based on an event, it contains rules that create a quick chain of events developed by the compound story M section.
  • Mental reaction knowledge base Contains rules for creating mental reactions that occur with the occurrence of events as external actions. Mental reactions are expressed as a concept of modification. For example, a “broken” event may result in a mental reaction such as “pain”, “painful”, or “discomfort J”.
  • the collection of concepts in the conceptual knowledge base constitutes a network of frames / instances / networks.
  • each concept has some properties that define its properties, but in many cases these properties are linked to other concepts.
  • the attributes of the event concept may describe the name of this concept and other ideas that have a specific M role, or the name of a rule in the narrative knowledge base . This makes it possible to directly search for other concepts that have a specific relationship with this creed, or indirectly obtain related concepts through the rules in the narrative knowledge base. In the latter case, other event concepts having some relationship with the original event concept and the sequence thereof are calculated on condition that the original event concept has a value.
  • the nature of the M family in such event concepts is used in the definition of narrative techniques.
  • the narrative branch method is defined as an I »! Number that returns an event or tree with one or more elephants or narrative parts of the narrative as arguments.
  • many narrative techniques perform calculations by using the link M between components in the knowledge base.
  • Each component in the conceptual knowledge base is represented in the form of a frame with multiple attributes-values ⁇ , and the attribute value is often the other component in the conceptual knowledge base * or a specific rule in the narrative knowledge base.
  • the main type of the object 1 ⁇ branch method is to create a new output W report by taking this link and searching for other concepts that have a specific relation to the argument.
  • the argument of the narrative technique in the narrative creation cycle is the application point of the particular narrative technique in the narrative tale, which is the current sweater, and this node is disguised as the inclination given by the narrative technique.
  • the whole story tree is expanded or transformed.
  • . 1 was a specific event an argument in ⁇ it ⁇ line operations to O..: 2. Take a specific subtree in the story tree as an argument and operate on it.
  • 3 is the case of the “synthesis” group of the narrative branch method
  • 4 is the case of adding and updating the character of a specific narrative forefact.
  • the difference between 1 and 2 is that the argument is only a single event or a partial bold, and from the difference in the way bold is made. Both can be divided into the following six types. The pattern is shown in Fig. 13 to Fig. 18 and that of type 2 is shown in Fig. 9 to Fig. 22
  • Containment type Returns a subtree containing the event or subtree of the argument as part of a larger tree.
  • Story technique 1 "main S of: inclusion", ⁇ eye building - Planning: inclusion ", r script: inclusion” such as corresponds to this.
  • the narrative branch method “Script: Inclusion” retrieves a script that contains the event of the argument in the script knowledge base in the narrative knowledge base, and outputs this script.
  • Kasuga replacement type One of the elements of the Kasuga required to be replaced with another. For example, instantiation of a frame during an event, instantiation of an instance, implementation of the concept of an event, and abstraction.
  • the narrative techniques “embodiment”, “abstraction”, “case”, “classification j”, etc. are equivalent to this.
  • the story branch method “case J” is any component of the argument event. With its instance.
  • Argument substitution type Replaces a newly created event or subtree with an argument event or subtree.
  • Narrative techniques such as “script: development j” and “goal-plan: development” correspond to this. For example, by referring to the rules in the knowledge base, which is the value of the narrative branch method “Script: Evolution j, the slot of the concept of the event in the argument event, the script,” Creates a first-speed event chain that composes the script for the event, and replaces it with the event of the argument.
  • Insertion type Insert a newly created event or subtree in the middle of the argument part.
  • the story-branch method import J group corresponds to this.
  • the narrative technique rKase J inserts one or more events in between the events given as arguments to prevent the arrest of subsequent events.
  • Modification type Modifies the ellipse at the end of the argument, that is, the way events are combined.
  • the narrative branch method r deformation from the viewpoint "corresponds to this.
  • the story technique r deformation from viewpoint '' is to rearrange the structure of this capital based on the viewpoint of a specific character appearing in the argument.
  • Desire-satisfaction An event of an argument is regarded as a hope, an event sequence for satisfying it is generated, and both are connected by a really-satisfaction relation. Similarity) ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ ⁇ )))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))
  • Problem-Solution Generates a sequence of events to solve the problem of the argument and solves the problem. Similar to the goal-planning (deployment) method. > Interrogation-Solution (Inclusion): A problem in which the event or subtree of the argument is included in the sequence of events for solving the problem. Is included in this sequence of events. (Similar to the goal-total (include) method.)
  • Prohibition-Violation Searches the knowledge base for a sequence of prohibited ones that includes the argument event or subtree in the sequence of violations against prohibition, and searches the event or subtree of the argument in this event sequence. (Similar to the Goal-Plan (Inclusion) method.)
  • Instruction-Compliance An instruction that includes the event or subtree of the argument in the event sequence for complying with the instruction.
  • the event sequence of compliance is searched from the knowledge base, and the event or subtree of the argument is searched for this event. Include in a row. (Similar to the Goal-Plan (Inclusion) method.)
  • Damage-elimination Search the knowledge base for the damage-elimination event sequence that includes the argument event or subtree in the event sequence for eliminating the damage, and find the argument event or subtree as this event. Include in a row. (Similar to the goal-total (include) method.)
  • a second event or subtree that occurs as a reaction to the first event or subtree is connected by a reaction relationship.
  • Recollection Creates a sequence of recollected past events for any object in the current situation. aThe event sequence must have actually occurred.
  • Movement Transform a story structure by moving a specific event or subtree in a subtree.
  • the event or source of the move is marked with an “omitted” symbol e
  • EVENT Xte tribeishai! '(AGT AGT &) (TIM T1M & Time 1) (LOC L0C & Location!>): EVENT); This key: EVENT specifies the contrast of the event concept. In addition, it is possible to specify the character and the object;
  • Target ($ W Section-Purpose (Want ((AGT AGT & X) (TIM NIL) (LOC NIL)))
  • the argument is the inclusion form of the event.
  • Previous Is defined as a set of rules for determining the nodes in the story tree to which the narrative technique is to be applied next, and the latter is defined as a set of rules for determining the accepted technique that should actually apply to that node. Then
  • story wartime rules may be referred to as story technique application point decision strategy rules and story technique decision strategies rules, respectively.
  • Each rail has in its conditional part story parameters and other information that define the characteristics of the story to be created, and a conclusion about the application points or story branching methods of the candidate story techniques.
  • the story parameters are basically given to the system by the user at the beginning of ft, but the value of the story parameters is set to a global variable, and the system refers to the story parameters in each cycle of story creation, so the value can be changed at any time. Even if it is changed, it is possible to operate and train as it is based on the changed story parameters. Therefore, the story parameters can be arbitrarily changed during the creation of the material.
  • the system has the ability to create stories with different structures at any time.
  • the terminal node is set as the application point # from left to right, and the same process is repeated for the terminal node of the newly added subtree.
  • Each rule consists of a condition part and a conclusion part.
  • the story parameter * story type * specifies the type of story, such as "novel”, and includes the story of advertisement j and "scenario".
  • the M number check-gp-type determines the type of application point of the parent-parent technique, that is, whether the terminal node is an event or the other nodes are subtrees.
  • the W-number check-gp-level checks the position of the current application point of the story technique in the story tree as a level. 0 corresponds to the,, / pre-level node, and one less is added.
  • the M-number set-gp-const used in the conclusion describes the constraints on nodes that can be applied to the story technique in the subsequent expansion of the story tree.
  • * frowth- list * is e example is ⁇ one eye of the rules are stored or story technique-class point as a argument in the , * Story of eve * power r novel j. Place person-intention * value is 1 "high”.
  • Task is r story creation ", argument type is arbitrary and S and L bells are"
  • Story type * "Specify the type of story, such as novel j or ⁇ scenario j. Such a story type creation is mainly based on a framework that can be realized through the way of constructing story narratives. Has become. 2.
  • ma * Instruct the use of a specific pattern provided in the system. For example, a theme pattern such as " ⁇ Murder by inheritance J.” ing.
  • Length Specify the length of story, professional, and / or video. This is currently determined by the number of events contained in each, and the relationship between length and structure has not yet been specifically examined.
  • Character-Intention * Specify the color of the character's intentional action that affects the structure of the story. This is a strong stand, and the story tends to be developed by the planning action based on the character's goal.
  • Point of View Specify the viewpoint to tell the story.
  • the types of viewpoints can be broadly classified into omnidirectional viewpoints and partial viewpoints.
  • the former is a kind of divine point of view that knows all the characters and situations in the story.
  • the latter is the intimate point of view of a particular character. This is especially effective at reorganizing the once created narrative content into narrative expressions.
  • events that a particular character does not know directly need to be indirectly introduced into the story by communication based on discourse or recollection.
  • Repeatability * Specifies the degree to which an event or its repetition, characters with similar characteristics, etc. are repeated.
  • the system proceeds with story creation by repeating the cycle of determining the point of passing on the story technique, determining and executing the story branch method, and expanding or changing the story tree as a whole.
  • Figure 23 shows such a
  • the determination and execution of the narrative technique is performed by the following procedure I ⁇ .
  • Rule that determines the condition part of the narrative technique determination rules Collect all the rules that can be passed You.
  • the rule for determining the narrative technique is as follows: the condition part is the current face ', the value of the narrative parameter, the type of applicable narrative technique application point (event or subtree), the position of the narrative technique application point in the narrative tree (story ⁇ , Etc.) are described.
  • the user sets the importance of each story parameter in advance, and the addition of this importance determines the priority of the applicable rule. Performed on application points. This is one cycle of executing the language technique. Note that the priority of a rule once executed is decremented at a fixed rate. At the end of this, a new cycle starts with the collection of available rules again until the termination conditions in each phase are satisfied.
  • the end conditions are determined by applying one story technique for the creation of a core story, creating a professional story, and creating a story, and judging the story creation by the value of the story parameter * length *. I have.
  • the narrative technique is given the power to be given an event or subtree as an argument, as described above, which corresponds to any node or part in the narrative tree.
  • the resulting event or subtree returned by the end of the story replaces this part of the argument, thereby forming the story as a whole.
  • the story tree is expanded or transformed by replacing the new subtree created by applying the story technique to the points where the story branch method is applied in the story tree with the points for the compost. This replacement is done simply Te L isp W the subst of l3 ⁇ 4 was use Rere following structure-expander (Koyo' ⁇
  • a story tree is constructed at several levels from the target story text, and the characteristics of the story tree and the relationships used in the story tree are described with reference to the various backgrounds described above. organize.
  • the process of creating the story can be assumed as a process to reverse this, and at this time, it is shown that the story branch method as described above is used.
  • the subject of analysis is short story fiction f by Maisa Yamamura f Maiko Murder Case J.
  • the following analysis is performed along the following hand jet. Increasingly, the sentences that make up the superficial text of the story are divided into phrase levels, and each phrase becomes a single elephant, and the entire e WIS that composes a story based on the speed of events is swiftly. It is difficult to bundle a single story.
  • the word tree is basically made up of "scenes" that follow. Next, each scene is considered as one event, and the whole structure of the story is formed into one tree. The above is equivalent to the pro / exposure analysis.
  • the events in the narrative composed here are not necessarily arranged according to the order of time in the narrative world. A new story tree is constructed by restoring the engagement. This story tree corresponds to the story level.
  • a scene is a group of semantic units in a narrative determined by the same characters, timeliness and location. Therefore, a plurality of such story-unit story trees are configured. This is called an “r scene tree”.
  • the level of the fast chain of events at the terminal node in the scene tree is called "construction”. Since a story is a unique form of discourse that revolves around a chain of events based on a time cabinet, it is useful to introduce such a unit of scene defined by time and space. (However, there is no criterion for the continuity of the time and the identity of the place. For example, some scenes are converted at the second level, while others are longer-term.
  • Figure 24 shows an example in which the following S-head scene is represented by a story tree.
  • Lips are pale pink, like a peach blossom.
  • Tables 6 to 8 summarize the narratives composed of scenes as basic units. When a scene cannot be represented by one event, the scene is represented as a succession of several events. In these tables, this is indicated as “Scene ⁇ Event”.
  • pro pattern As a set of nodes below the vertices in the story tree, leading to a chain of events that correspond to the scene or the sub-events needed to create the scene.
  • the thing ⁇ tree is called “pro” / to tree j.
  • the story summarized at the level of the terminal node of the professional ', / tota, that is, at the level of the scene or its lower event, is referred to as "pro', / to j. Is divided into Figure 25 and Figure 35.
  • Embedded scenes may not always be real (for a story), but they may be fictional scenes, false guesses, predictions / estimations, etc. (a kind of possible world).
  • the temporal non-linearity of the development of such cases is most pronounced in reasoning novels, or similar structures are frequently found in many other stories.
  • the character may go back to the childhood reminiscence, or in some parts of the story, past episodes may be recollected as flashba's.
  • flashba's The reader understands the story as a whole by restoring these intricate times into a series of lines.
  • the mystery novel When the culprit and his motive are known, the mystery is resolved when the story is restored along a single time axis, and the story ends.
  • the narrative that evolves along the time axis that is, the narrative content, is called “ r story for a program”. Therefore, the structure of the narrative content must then be clarified by restoring the story as another structural level based on professionals.
  • the structure of the story can be restored by rearranging the end of the pro ',.'
  • the professional tree is governed by an abstract knowledge unit of professional, y-patterns.
  • the story's story is called “story.
  • a story is (This refers to the chain of terminal nodes in the story tree.)
  • ⁇ is considered to be controlled by a knowledge unit called theme pattern j.
  • theme pattern j In the case of this example.
  • Dg Property Acquisition Assuming that one of the bars is used, it is assumed that the heir is used. The emergence of-genuine killing (losing) "and ⁇ investigation, reasoning-arrest (success) J are interposed between them (and the introduction of the detective role is added).
  • This puma pattern is a different concept from the above-mentioned professional pattern, and the same story created using a certain theme pattern can be expressed by various different plot patterns.
  • the plot tree and the story tree are governed by different principles, as shown in Tables 9 to 12.
  • Table 9 to Table 12 show the fast chain of the scene or its constituent elements.
  • the plots from Table 6 to Table 8 were transformed into a story. 2 means the embedding scene, and the number in the brackets "/" also means the lower event.
  • For example--46 25-1-1 indicates the embedded scene 1 of the scene 25 in the professional game.
  • Table 9 Power and Tables 12 show that professionals, as a representation of objects, are organized according to the order of narrative of the time mechanism in the narrative world. . This is shown in Fig. 36 through Fig. 40, which is a story story structured based on the theme pattern.
  • Fig. 41 shows the relationship between the professional tree and the tree; the tree has the following responsive relationship.
  • this theme pattern is dazzling in such a way that the development centered on the murderer (Niro Sugino) and the development centered on the detective roles (Kogiku, Sawaki) are mixed. I get stuck.
  • the * marked above is the development of Murderer II.
  • the pattern of “pro” is based on the development centered on the detective role. «!
  • the development of the detectives in the theme patterns is almost the same as each « component in the professional ',' pattern, that is, 'Detective appearance introduction, murder
  • the "make up of false heirs" in the poma pattern is a professional, through a speech based on Kogiku's recollections and inferences in “investigation and reasoning (2)” in the professional pattern, (25-1- ⁇ 25-1-5, 28-1).
  • "Genuine appearance” is based on Kogiku's inference in "key S.
  • Creating a story from a theme is an extension of the theme pattern that expresses the theme to a story tree.
  • This section describes the structural transformation from one story tree to a plot tree, which is the end of the story for expressing the story.
  • the input to the European system is one or more fragmentary events, but here we have an example (kill ((agt agt &-ro) (ob.iagt & small)) (tim tim & midnight 1) (loc loc & Place 1))) as input.
  • An event concept such as "kill” is put at the top, followed by a number of character pairs such as subject (agt), object (0 bj), time (tim), and location (loc>) and their value pairs.
  • the list is called an “event.”
  • Each value is the name of a component in the knowledge base, and the & symbol is used in the value to indicate an instance in the knowledge base.
  • agt &-ro, agt & small, tim & midnight 1, lac & location 1 are instances, whereas in the case of a frame, the 3 ⁇ 4 symbol is like an agtY dancer.
  • Each instance contains some specific characteristics: "Ichiro kills Kozuzu at midnight 1 at location 1 J .
  • the core story is a form of ft.'S first story tree format that converts a fragmentary event of input into a story tree format.
  • the object is a hierarchical tree with the event as the terminal node and the cabinet member as the other node.
  • the name of the M department is preceded by a $ sign. This is illustrated in FIG.
  • the fast starvation of events in this story tree is the core story. this is,
  • Hayakawa made Kozuki. Long wanted to monopolize Hayakawa's property.
  • Ichiro killed Kozuki. Ichiro robbed Kozushi of money. Sava reveals who killed Kozushi.
  • Police arrested Ichiro. Police confirmed Ichiro.
  • the narrative technique to be executed at each point in the creation is determined through a flurry of narrative strategies that refer to narrative parameters such as.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

Un système de support destiné à l'automatisation de la préparation de structures d'histoires prépare d'une manière automatique une structure d'histoire fondamentalement dans la séquence d'une histoire centrale, d'une histoire et d'une trame sur la base d'événements fragmentaires fournis par un utilisateur ou un arbre partiel, ainsi que les valeurs de paramètres d'histoire. La structure d'histoire est constituée uniformément par un système d'arbre et le système comprend des paramètres d'histoire sous la forme de groupes de modalités fragmentaires permettant d'élargir/modifier cet arbre, une base de connaissance contenant une connaissance relative à des concepts ainsi qu'à une combinaison d'événements auxquels le système fait référence, une stratégie d'histoire comprenant des groupes de règles fragmentaires destinés à gérer le fonctionnement des groupes de techniques d'histoire ainsi qu'un mécanisme de gestion de préparation d'histoire destiné à gérer tous ces processus. Différentes histoires peuvent être préparées de manière souple à partir du même événement d'entrée ou d'un même arbre partiel en fonction du type des paramètres de l'histoire entrés par l'utilisateur, et cette capacité est offerte du fait que tous les mécanismes principaux destinés à la préparation de la structure d'histoire, tels que la technique d'histoire, la stratégie d'hoistoire et la base de connaissance comprennent les informations fragmentaires regroupées, et du fait que le système met en oeuvre des combinaisons polyvalentes à l'aide du procédé cohérent ci-dessus décrits.
PCT/JP1996/000813 1995-03-28 1996-03-28 Systeme de support destine a l'automatisation de la preparation de structures d'histoires WO1996030844A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP9322595 1995-03-28
JP7/93225 1995-03-28

Publications (1)

Publication Number Publication Date
WO1996030844A1 true WO1996030844A1 (fr) 1996-10-03

Family

ID=14076613

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP1996/000813 WO1996030844A1 (fr) 1995-03-28 1996-03-28 Systeme de support destine a l'automatisation de la preparation de structures d'histoires

Country Status (1)

Country Link
WO (1) WO1996030844A1 (fr)

Cited By (30)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2001033479A1 (fr) * 1999-11-04 2001-05-10 Matsushita Electric Industrial Co., Ltd. Procede et appareil d'assistance a la creation et support enregistre
JP2009522691A (ja) * 2006-01-03 2009-06-11 イーストマン コダック カンパニー 情報伝達物の作成方法
WO2011126498A1 (fr) * 2010-04-09 2011-10-13 Microsoft Corporation Création automatisée d'histoires
US10572606B1 (en) * 2017-02-17 2020-02-25 Narrative Science Inc. Applied artificial intelligence technology for runtime computation of story outlines to support natural language generation (NLG)
US10657201B1 (en) 2011-01-07 2020-05-19 Narrative Science Inc. Configurable and portable system for generating narratives
US10699079B1 (en) 2017-02-17 2020-06-30 Narrative Science Inc. Applied artificial intelligence technology for narrative generation based on analysis communication goals
US10706236B1 (en) 2018-06-28 2020-07-07 Narrative Science Inc. Applied artificial intelligence technology for using natural language processing and concept expression templates to train a natural language generation system
US10747823B1 (en) 2014-10-22 2020-08-18 Narrative Science Inc. Interactive and conversational data exploration
US10755042B2 (en) 2011-01-07 2020-08-25 Narrative Science Inc. Automatic generation of narratives from data using communication goals and narrative analytics
US10853583B1 (en) 2016-08-31 2020-12-01 Narrative Science Inc. Applied artificial intelligence technology for selective control over narrative generation from visualizations of data
CN112035654A (zh) * 2020-09-01 2020-12-04 上海松鼠课堂人工智能科技有限公司 题目自动衍生生成方法
US10943069B1 (en) 2017-02-17 2021-03-09 Narrative Science Inc. Applied artificial intelligence technology for narrative generation based on a conditional outcome framework
US10963649B1 (en) 2018-01-17 2021-03-30 Narrative Science Inc. Applied artificial intelligence technology for narrative generation using an invocable analysis service and configuration-driven analytics
US10990767B1 (en) 2019-01-28 2021-04-27 Narrative Science Inc. Applied artificial intelligence technology for adaptive natural language understanding
US11042708B1 (en) 2018-01-02 2021-06-22 Narrative Science Inc. Context saliency-based deictic parser for natural language generation
CN113033182A (zh) * 2021-03-25 2021-06-25 网易（杭州）网络有限公司 文本创作的辅助方法、装置及服务器
US11068661B1 (en) 2017-02-17 2021-07-20 Narrative Science Inc. Applied artificial intelligence technology for narrative generation based on smart attributes
US11170038B1 (en) 2015-11-02 2021-11-09 Narrative Science Inc. Applied artificial intelligence technology for using narrative analytics to automatically generate narratives from multiple visualizations
US11222184B1 (en) 2015-11-02 2022-01-11 Narrative Science Inc. Applied artificial intelligence technology for using narrative analytics to automatically generate narratives from bar charts
US11232268B1 (en) 2015-11-02 2022-01-25 Narrative Science Inc. Applied artificial intelligence technology for using narrative analytics to automatically generate narratives from line charts
US11238090B1 (en) 2015-11-02 2022-02-01 Narrative Science Inc. Applied artificial intelligence technology for using narrative analytics to automatically generate narratives from visualization data
US11288328B2 (en) 2014-10-22 2022-03-29 Narrative Science Inc. Interactive and conversational data exploration
US11521079B2 (en) 2010-05-13 2022-12-06 Narrative Science Inc. Method and apparatus for triggering the automatic generation of narratives
US11561684B1 (en) 2013-03-15 2023-01-24 Narrative Science Inc. Method and system for configuring automatic generation of narratives from data
US11568148B1 (en) 2017-02-17 2023-01-31 Narrative Science Inc. Applied artificial intelligence technology for narrative generation based on explanation communication goals
KR20240000291A (ko) * 2022-06-23 2024-01-02 한국전자기술연구원 범죄사실 스토리 타임라인 기반 범죄사실 경과 분석 시스템 및 방법
US11922344B2 (en) 2014-10-22 2024-03-05 Narrative Science Llc Automatic generation of narratives from data using communication goals and narrative analytics
US11954445B2 (en) 2017-02-17 2024-04-09 Narrative Science Llc Applied artificial intelligence technology for narrative generation based on explanation communication goals
US11989659B2 (en) 2010-05-13 2024-05-21 Salesforce, Inc. Method and apparatus for triggering the automatic generation of narratives
US12153618B2 (en) 2015-11-02 2024-11-26 Salesforce, Inc. Applied artificial intelligence technology for automatically generating narratives from visualization data

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
FILES BY THE KNOWLEDGE BASE SYSTEM STUDY SESSION OF THE JAPANESE SOCIETY FOR ARTIFICIAL INTELLIGENCE, Vol. 19, (1992), TAKASHI OGATA, TAKAO TERAO, "Particularization of Outline in Story Generating System Using EBL", pages 117-124. *
PAPERS FROM THE SYSTEM SYMPOSIUM, Vol. 19, (1993), TAKASHI OGATA, KOICHI WATANABE, KOICHI HORI, SETSUO OSUGA, "Application of Story Generating System to Life-Style Marketing", pages 217-224. *

Cited By (61)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2001033479A1 (fr) * 1999-11-04 2001-05-10 Matsushita Electric Industrial Co., Ltd. Procede et appareil d'assistance a la creation et support enregistre
JP5138852B2 (ja) * 1999-11-04 2013-02-06 株式会社こんてんつ 創作物作成支援装置
JP2009522691A (ja) * 2006-01-03 2009-06-11 イーストマン コダック カンパニー 情報伝達物の作成方法
WO2011126498A1 (fr) * 2010-04-09 2011-10-13 Microsoft Corporation Création automatisée d'histoires
US8422852B2 (en) 2010-04-09 2013-04-16 Microsoft Corporation Automated story generation
US9161007B2 (en) 2010-04-09 2015-10-13 Microsoft Technology Licensing, Llc Automated story generation
US11521079B2 (en) 2010-05-13 2022-12-06 Narrative Science Inc. Method and apparatus for triggering the automatic generation of narratives
US11989659B2 (en) 2010-05-13 2024-05-21 Salesforce, Inc. Method and apparatus for triggering the automatic generation of narratives
US10657201B1 (en) 2011-01-07 2020-05-19 Narrative Science Inc. Configurable and portable system for generating narratives
US11790164B2 (en) 2011-01-07 2023-10-17 Narrative Science Inc. Configurable and portable system for generating narratives
US11501220B2 (en) 2011-01-07 2022-11-15 Narrative Science Inc. Automatic generation of narratives from data using communication goals and narrative analytics
US10755042B2 (en) 2011-01-07 2020-08-25 Narrative Science Inc. Automatic generation of narratives from data using communication goals and narrative analytics
US11561684B1 (en) 2013-03-15 2023-01-24 Narrative Science Inc. Method and system for configuring automatic generation of narratives from data
US11921985B2 (en) 2013-03-15 2024-03-05 Narrative Science Llc Method and system for configuring automatic generation of narratives from data
US11922344B2 (en) 2014-10-22 2024-03-05 Narrative Science Llc Automatic generation of narratives from data using communication goals and narrative analytics
US10747823B1 (en) 2014-10-22 2020-08-18 Narrative Science Inc. Interactive and conversational data exploration
US11475076B2 (en) 2014-10-22 2022-10-18 Narrative Science Inc. Interactive and conversational data exploration
US11288328B2 (en) 2014-10-22 2022-03-29 Narrative Science Inc. Interactive and conversational data exploration
US11188588B1 (en) 2015-11-02 2021-11-30 Narrative Science Inc. Applied artificial intelligence technology for using narrative analytics to interactively generate narratives from visualization data
US12153618B2 (en) 2015-11-02 2024-11-26 Salesforce, Inc. Applied artificial intelligence technology for automatically generating narratives from visualization data
US11238090B1 (en) 2015-11-02 2022-02-01 Narrative Science Inc. Applied artificial intelligence technology for using narrative analytics to automatically generate narratives from visualization data
US11232268B1 (en) 2015-11-02 2022-01-25 Narrative Science Inc. Applied artificial intelligence technology for using narrative analytics to automatically generate narratives from line charts
US11222184B1 (en) 2015-11-02 2022-01-11 Narrative Science Inc. Applied artificial intelligence technology for using narrative analytics to automatically generate narratives from bar charts
US11170038B1 (en) 2015-11-02 2021-11-09 Narrative Science Inc. Applied artificial intelligence technology for using narrative analytics to automatically generate narratives from multiple visualizations
US10853583B1 (en) 2016-08-31 2020-12-01 Narrative Science Inc. Applied artificial intelligence technology for selective control over narrative generation from visualizations of data
US11341338B1 (en) 2016-08-31 2022-05-24 Narrative Science Inc. Applied artificial intelligence technology for interactively using narrative analytics to focus and control visualizations of data
US11144838B1 (en) 2016-08-31 2021-10-12 Narrative Science Inc. Applied artificial intelligence technology for evaluating drivers of data presented in visualizations
US10762304B1 (en) 2017-02-17 2020-09-01 Narrative Science Applied artificial intelligence technology for performing natural language generation (NLG) using composable communication goals and ontologies to generate narrative stories
US12314674B2 (en) 2017-02-17 2025-05-27 Salesforce, Inc. Applied artificial intelligence technology for narrative generation based on a conditional outcome framework
US11954445B2 (en) 2017-02-17 2024-04-09 Narrative Science Llc Applied artificial intelligence technology for narrative generation based on explanation communication goals
US11568148B1 (en) 2017-02-17 2023-01-31 Narrative Science Inc. Applied artificial intelligence technology for narrative generation based on explanation communication goals
US11983503B2 (en) 2017-02-17 2024-05-14 Salesforce, Inc. Applied artificial intelligence technology for narrative generation based on a conditional outcome framework
US11562146B2 (en) 2017-02-17 2023-01-24 Narrative Science Inc. Applied artificial intelligence technology for narrative generation based on a conditional outcome framework
US10943069B1 (en) 2017-02-17 2021-03-09 Narrative Science Inc. Applied artificial intelligence technology for narrative generation based on a conditional outcome framework
US10572606B1 (en) * 2017-02-17 2020-02-25 Narrative Science Inc. Applied artificial intelligence technology for runtime computation of story outlines to support natural language generation (NLG)
US10699079B1 (en) 2017-02-17 2020-06-30 Narrative Science Inc. Applied artificial intelligence technology for narrative generation based on analysis communication goals
US12086562B2 (en) 2017-02-17 2024-09-10 Salesforce, Inc. Applied artificial intelligence technology for performing natural language generation (NLG) using composable communication goals and ontologies to generate narrative stories
US11068661B1 (en) 2017-02-17 2021-07-20 Narrative Science Inc. Applied artificial intelligence technology for narrative generation based on smart attributes
US12423525B2 (en) 2017-02-17 2025-09-23 Salesforce, Inc. Applied artificial intelligence technology for narrative generation based on explanation communication goals
US11816438B2 (en) 2018-01-02 2023-11-14 Narrative Science Inc. Context saliency-based deictic parser for natural language processing
US11042708B1 (en) 2018-01-02 2021-06-22 Narrative Science Inc. Context saliency-based deictic parser for natural language generation
US11042709B1 (en) 2018-01-02 2021-06-22 Narrative Science Inc. Context saliency-based deictic parser for natural language processing
US10963649B1 (en) 2018-01-17 2021-03-30 Narrative Science Inc. Applied artificial intelligence technology for narrative generation using an invocable analysis service and configuration-driven analytics
US11561986B1 (en) 2018-01-17 2023-01-24 Narrative Science Inc. Applied artificial intelligence technology for narrative generation using an invocable analysis service
US11003866B1 (en) 2018-01-17 2021-05-11 Narrative Science Inc. Applied artificial intelligence technology for narrative generation using an invocable analysis service and data re-organization
US12001807B2 (en) 2018-01-17 2024-06-04 Salesforce, Inc. Applied artificial intelligence technology for narrative generation using an invocable analysis service
US11023689B1 (en) 2018-01-17 2021-06-01 Narrative Science Inc. Applied artificial intelligence technology for narrative generation using an invocable analysis service with analysis libraries
US11334726B1 (en) 2018-06-28 2022-05-17 Narrative Science Inc. Applied artificial intelligence technology for using natural language processing to train a natural language generation system with respect to date and number textual features
US10706236B1 (en) 2018-06-28 2020-07-07 Narrative Science Inc. Applied artificial intelligence technology for using natural language processing and concept expression templates to train a natural language generation system
US11042713B1 (en) 2018-06-28 2021-06-22 Narrative Scienc Inc. Applied artificial intelligence technology for using natural language processing to train a natural language generation system
US11989519B2 (en) 2018-06-28 2024-05-21 Salesforce, Inc. Applied artificial intelligence technology for using natural language processing and concept expression templates to train a natural language generation system
US11232270B1 (en) 2018-06-28 2022-01-25 Narrative Science Inc. Applied artificial intelligence technology for using natural language processing to train a natural language generation system with respect to numeric style features
US10990767B1 (en) 2019-01-28 2021-04-27 Narrative Science Inc. Applied artificial intelligence technology for adaptive natural language understanding
US12288039B1 (en) 2019-01-28 2025-04-29 Salesforce, Inc. Applied artificial intelligence technology for adaptively classifying sentences based on the concepts they express to improve natural language understanding
US11341330B1 (en) 2019-01-28 2022-05-24 Narrative Science Inc. Applied artificial intelligence technology for adaptive natural language understanding with term discovery
CN112035654B (zh) * 2020-09-01 2025-04-22 上海松鼠课堂人工智能科技有限公司 题目自动衍生生成方法
CN112035654A (zh) * 2020-09-01 2020-12-04 上海松鼠课堂人工智能科技有限公司 题目自动衍生生成方法
CN113033182B (zh) * 2021-03-25 2024-03-01 网易（杭州）网络有限公司 文本创作的辅助方法、装置及服务器
CN113033182A (zh) * 2021-03-25 2021-06-25 网易（杭州）网络有限公司 文本创作的辅助方法、装置及服务器
KR102735437B1 (ko) 2022-06-23 2024-11-29 한국전자기술연구원 범죄사실 스토리 타임라인 기반 범죄사실 경과 분석 시스템 및 방법
KR20240000291A (ko) * 2022-06-23 2024-01-02 한국전자기술연구원 범죄사실 스토리 타임라인 기반 범죄사실 경과 분석 시스템 및 방법

Similar Documents

Publication Publication Date Title
WO1996030844A1 (fr) Systeme de support destine a l'automatisation de la preparation de structures d'histoires
Parikh et al. Indian Women (an inner dialogue)
Lunenfeld The digital dialectic: New essays on new media
Heffernan Post-apocalyptic culture: Modernism, postmodernism, and the twentieth-century novel
Vizenor Narrative chance: postmodern discourse on native American Indian literatures
Spariosu Modernism and exile: Liminality and the utopian imagination
Swenson Society, spirituality, and the sacred: A social scientific introduction
Swirski From literature to biterature: Lem, Turing, Darwin, and explorations in computer literature, philosophy of mind, and cultural evolution
Hamilton The World of Failing Machines: Speculative Realism and Literature
Gill Creative encounters, appreciating difference: perspectives and strategies
Sorlin A Linguistic Approach to David Mitchell's Science-Fiction Stories in Cloud Atlas
Taylor The Foundations of Nature: Metaphysics of gift for an integral ecological ethic
Prophet Evolution Esotericized: Conceptual Blending and the Emergence of Secular, Therapeutic Salvation
Hughes Graced Life
Handley et al. Don't Teach Coding: Until You Read This Book
Rico Imagining Global Female Futures in Black Speculative and Science Fiction
Margolies et al. Utopia and Modernity in China: Contradictions in Transition
O'Farrell A memorial in the world: legendary patterns in Late Antique biography
Fawcett Changeling the Outcome: Fairy-Tale Framing and the Ethics of Care in Michelle Lovretta's Killjoys
ÒGÚNSÀNYÀ MUSICAL MOTIFS IN THE DRAMATIC WORKS OF WỌLÉ SÓYÍNKÁ AND FẸ́MI ÒṢ ÓFISAN
Brown Shakespearean Spiders and Shakespearean Slime Mold: Allusion and Posthuman Identity in Adrian Tchaikovsky’s Children of Time Novels
Korda ‘LIBRETTISING’SCIENCE: Using operatic narrative and performance to re-present scientific thought in an investigation of new methods towards developing contemporary opera.
Wear The Cost of Work: Performance, Asian Diasporas, and Neoliberalism
GRISELDA et al. TRANSLATION OF ALLEGORY OR ALLEGORY OF TRANSLATION? PETRARCH'S REDRESSING
Subasinghe When Kōlam Rebecomes: A Performative Exploration of Form, Philosophy and Contemporary Relevance of a Sinhalese Traditional Performance Practice

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): CN JP US

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
122 Ep: pct application non-entry in european phase