CN110209908A - Application recommended method, device, computer equipment and computer storage medium based on user interest portrait - Google Patents
Application recommended method, device, computer equipment and computer storage medium based on user interest portrait Download PDFInfo
- Publication number
- CN110209908A CN110209908A CN201910319420.9A CN201910319420A CN110209908A CN 110209908 A CN110209908 A CN 110209908A CN 201910319420 A CN201910319420 A CN 201910319420A CN 110209908 A CN110209908 A CN 110209908A
- Authority
- CN
- China
- Prior art keywords
- word
- user
- interest
- notional
- father
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/958—Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Document Processing Apparatus (AREA)
Abstract
This application discloses a kind of application recommended methods and device based on user interest portrait, are related to data analysis technique field, can construct more reasonable user interest portrait, improve the accuracy for recommending user's application based on user interest portrait.The described method includes: obtaining the Ontological concept word for reflecting each point of interest, and the web page text word that user browses webpage behavior is mapped to the Ontological concept word for being used to reflect each point of interest, obtains interest value of the user on each Ontological concept word；Interest value of the user on each Ontological concept word is updated according to the hierarchical relationship between each Ontological concept word, obtains effective interest value of the user on each Ontological concept word；Effective interest value based on user on each Ontological concept word, building record the user interest portrait of each point of interest；According to effective interest value of each point of interest, by effective interest value ranking before default value and application relevant to point of interest is recommended to user.
Description
Technical field
The present invention relates to data analysis technique fields, apply recommendation side particularly with regard to what is drawn a portrait based on user interest
Method, device, computer equipment and computer storage medium.
Background technique
With the development of big data, the explosive growth of data volume and the maturation of big data analysis technology catch user
The behavioral data caught is more and more, and user's portrait is also more and more abundant with the surge of information, more and more finely, can be applied
Into client's marketing of certain industries itself.
The behavior that webpage is browsed by collecting user can construct user interest portrait, and user interest portrait includes more
A point of interest label and user are to the fancy grade of each point of interest, for example, user is 60, user to the fancy grade of sport
Fancy grade to music is 20 etc..The prior art would generally preset multiple points of interest, when in website generate user it is clear
Look at behavior when, by user browse web sites in web page text be mapped on the Ontological concept for characterizing different points of interest, such as will be sufficient
Ball is mapped to the points of interest such as ball, sport, thus construct user interest portrait, and based on user interest draw a portrait to user recommend it is emerging
Interest is worth higher application.
However, the prior art is during constructing user interest portrait, between the Ontological concept for characterizing point of interest
There may be hierarchical relationship, for example, shuttlecock and basketball belong to it is ball, if by web page text mapping show shuttlecock and
The interest value of basketball is higher, and lower to ball interest value, so that user interest is drawn a portrait, building there is a situation where unreasonable, lead
Cause the application inaccuracy that user is recommended based on user interest portrait.
Summary of the invention
In view of this, the present invention provides a kind of application recommended method based on user interest portrait, device, computers to set
Standby and computer storage medium, main purpose are to solve to draw a portrait based on user interest to recommend the asking using inaccuracy of user
Topic.
According to the present invention on one side, a kind of application recommended method based on user interest portrait, this method packet are provided
It includes:
The Ontological concept word for reflecting each point of interest is obtained, and the web page text word that user browses webpage behavior is reflected
It is mapped to the Ontological concept word for reflecting each point of interest, obtains interest value of the user on each Ontological concept word；
Interest value according to the hierarchical relationship between each Ontological concept word to the user on each Ontological concept word
It is updated, obtains effective interest value of the user on each Ontological concept word；
Effective interest value based on the user on each Ontological concept word, the user that building records each point of interest are emerging
Interest portrait；
According to the effective interest value for recording each point of interest in user interest portrait, by effective interest value ranking
Before default value and application relevant to point of interest is recommended to user.
Further, general in each ontology to the user in the hierarchical relationship according between each Ontological concept word
The interest value read on word is updated, and obtains user before the update interest value on each Ontological concept word, the method is also
Include:
Ontological concept word is divided into father's notional word with hierarchical relationship and sub- notional word in advance, under each father's notional word
There are multiple sub- notional words.
Further, the hierarchical relationship according between each Ontological concept word is to the user in each Ontological concept
Interest value on word is updated, and is obtained effective interest value of the user on each Ontological concept word and is included:
According to the hierarchical relationship between each Ontological concept word, each sub- notional word for including under father's notional word is determined；
It is poor based on the level under father's notional word between each sub- notional word for including and father's notional word, calculate each height
Additional interest value of the notional word on father's notional word；
It is emerging on father's notional word that additional interest value of each sub- notional word on father's notional word is added to user
Interest value, effective interest value after being updated on each Ontological concept word.
Further, the level based under father's notional word between each sub- notional word for including and father's notional word
Difference, calculating additional interest value of each sub- notional word on father's notional word includes:
It is poor according to the level between each sub- notional word for including under father's notional word and father's notional word, determine each height
Level accounting coefficient of the notional word on father's notional word；
Level accounting coefficient and user of each sub- notional word on father's notional word are calculated separately in each sub- concept
The product between interest value on word obtains additional interest value of each sub- notional word on father's notional word.
Further, the level according between each sub- notional word for including under father's notional word and father's notional word
Difference determines that level accounting coefficient of each sub- notional word on father's notional word includes:
Poor according to the level between each sub- notional word for including under father's notional word and father's notional word, generation layer is differential
It is reciprocal；
It calculates the level difference inverse and presets the product in level between the level parameter of each sub- notional word, really
Fixed level accounting coefficient of each sub- notional word on father's notional word.
Further, in the level ginseng for calculating the level difference inverse and presetting each sub- notional word in level
Product between number, determine each sub- notional word before the level accounting coefficient on father's notional word, the method also includes:
According to the similarity in level between each sub- notional word and father's notional word, each sub- concept in the level is determined
The level parameter of word.
Further, the web page text word that user is browsed webpage behavior, which is mapped to, described is used to reflect each interest
On the Ontological concept word of point, obtaining interest value of the user on each Ontological concept word includes:
The web page text word that user browses webpage behavior is subjected to phase with the Ontological concept word for being used to reflect each point of interest
It is calculated like degree, obtains the similarity between web page text word and each Ontological concept word；
It is browsed after being greater than the similarity of preset threshold between cumulative web page text word and each Ontological concept word divided by user
Webpage quantity obtains interest value of the user on each Ontological concept word.
According to the present invention on the other hand, a kind of application recommendation apparatus based on user interest portrait, the dress are provided
It sets and includes:
Map unit browses webpage behavior for obtaining the Ontological concept word for reflecting each point of interest, and by user
Web page text word be mapped to described for reflecting the Ontological concept word of each point of interest, obtain user in each Ontological concept
Interest value on word；
Updating unit, for according to the hierarchical relationship between each Ontological concept word to the user in each Ontological concept
Interest value on word is updated, and obtains effective interest value of the user on each Ontological concept word；
Construction unit, for effective interest value based on the user on each Ontological concept word, building records each
The user interest of point of interest is drawn a portrait；
Recommendation unit will be described for recording effective interest value of each point of interest in drawing a portrait according to the user interest
Effectively interest value ranking is before default value and application relevant to point of interest is recommended to user.
Further, described device further include:
Division unit, in the hierarchical relationship according between each Ontological concept word to the user at each
Interest value on body notional word is updated, and obtains user before the update interest value on each Ontological concept word, in advance will
Ontological concept word is divided into father's notional word with hierarchical relationship and sub- notional word, and there are multiple sub- concepts under each father's notional word
Word.
Further, the updating unit includes:
Determining module, for according to the hierarchical relationship between each Ontological concept word, determining include under father's notional word each
A sub- notional word；
First computing module, for based under father's notional word between each sub- notional word for including and father's notional word
Level is poor, calculates additional interest value of each sub- notional word on father's notional word；
Adding module, for additional interest value of each sub- notional word on father's notional word to be added to user in father
Interest value on notional word, effective interest value after being updated on each Ontological concept word.
Further, first computing module includes；
Submodule is determined, for according to the layer between each sub- notional word for including under father's notional word and father's notional word
It is differential, determine level accounting coefficient of each sub- notional word on father's notional word；
Computational submodule, for calculating separately level accounting coefficient of each sub- notional word on father's notional word and using
Product of the family between the interest value on each sub- notional word obtains additional interest of each sub- notional word on father's notional word
Value.
Further, the determining submodule, specifically for according to each sub- notional word for including under father's notional word
Level between father's notional word is poor, the differential inverse of generation layer；
The determining submodule is specifically also used to calculate the level difference inverse and presets each sub- concept in level
Product between the level parameter of word determines level accounting coefficient of each sub- notional word on father's notional word.
Further, the determining submodule is specifically also used to calculate the level difference inverse described and preset
Product in level between the level parameter of each sub- notional word determines level accounting of each sub- notional word on father's notional word
Before coefficient, according to the similarity in level between each sub- notional word and father's notional word, determine that each height is general in the level
Read the level parameter of word.
Further, the map unit includes:
Second computing module, for user to be browsed to the web page text word of webpage behavior and is used to reflect each point of interest
Ontological concept word carries out similarity calculation, obtains the similarity between web page text word and each Ontological concept word；
Accumulator module, for add up between web page text word and each Ontological concept word greater than preset threshold similarity after
Webpage quantity is browsed divided by user, obtains interest value of the user on each Ontological concept word.
Another aspect according to the present invention provides a kind of computer equipment, including memory and processor, the storage
Device is stored with computer program, realizes that the application drawn a portrait based on user interest is pushed away when the processor executes the computer program
The step of recommending method.
Another aspect according to the present invention provides a kind of computer storage medium, is stored thereon with computer program, institute
The step of stating the application recommended method that realization is drawn a portrait based on user interest when computer program is executed by processor.
By above-mentioned technical proposal, the present invention provides a kind of application recommended method and device based on user interest portrait,
It is mapped to by obtaining the Ontological concept word for reflecting each point of interest, and by the web page text word that user browses webpage behavior
For reflecting the Ontological concept word of corresponding point of interest, interest value of the user on each Ontological concept word is obtained, according to each
Hierarchical relationship between Ontological concept word is updated interest value of the user on each Ontological concept word, obtains user each
Effective interest value on a Ontological concept word, the validity based on user on each Ontological concept word, building record each emerging
User's portrait of interest value, according to user interest draw a portrait in record effective interest value of each point of interest, related application is recommended
User.It is constructed to being directly based upon on the Ontological concept that the web page text of user's browsing is mapped to corresponding point of interest in the prior art
User interest portrait carry out the mode that user's application is recommended and compare, for characterizing interest in the user interest of prior art portrait
Between the Ontological concept of point and hierarchical relationship is not present, the present invention fully takes into account the layer between each Ontological concept word concept word
Grade relationship, is updated interest value of the user on father's notional word based on influence degree of the sub- notional word to father's notional word, from
And guarantee that the user interest portrait of building is more reasonable, improve the accuracy that user's application is recommended based on user interest portrait.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the present invention
Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Fig. 1 shows a kind of application recommended method process signal based on user interest portrait provided in an embodiment of the present invention
Figure；
Fig. 2 shows another application recommended method processes based on user interest portrait provided in an embodiment of the present invention to show
It is intended to；
Fig. 3 shows a kind of showing using the structure of recommendation apparatus based on user interest portrait provided in an embodiment of the present invention
It is intended to；
Fig. 4 shows another structure using recommendation apparatus based on user interest portrait provided in an embodiment of the present invention
Schematic diagram.
Specific embodiment
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
It is fully disclosed to those skilled in the art.
The embodiment of the invention provides a kind of application recommended methods based on user interest portrait, can construct more reasonably
User interest portrait improves the accuracy that user's application is recommended based on user interest portrait, as shown in Figure 1, this method comprises:
101, the Ontological concept word for reflecting each point of interest is obtained, and user is browsed to the web page text of webpage behavior
Word is mapped to the Ontological concept word for being used to reflect each point of interest, obtains interest of the user on each Ontological concept word
Value.
Wherein, Ontological concept word is the word that point of interest concept is indicated in this body surface, for example, amusement, politics, sport, sound
It is happy etc., it, specifically can be with word since each Ontological concept word has different characterizations, contextual feature etc. here without limiting
The each Ontological concept word of vectorial.
It should be noted that, although may include subclass relation in each Ontological concept word, for example, including shadow in amusement
Depending on, movement etc., football included below, basketball etc. are moved, is mutually juxtaposed between Ontological concept word here, each ontology is general
Word is read as individual term vector.
During browsing webpage, user can browse user for oneself interested webpage, for example, user compares
Like making up, then often can browse or search for makeups class webpage, user likes sport, then often can browse or search for sport
Class webpage, the process specifically searched for can be by jumping in other webpages, can also be and directly scanned for by search column.
Wherein, it is usually that user clicks the web page browsing behavior that the page generates, and passes through that user, which browses the behavior of webpage,
It clicks the different region of the page and jumps to the different display pages.When generating user's browsing webpage behavior, can specifically pass through
It is buried a little in front end or the daily record data of parsing rear end, the corresponding behavioral data of acquisition user's browsing webpage behavior is clear from user
It lookes in the corresponding behavioral data of webpage behavior and extracts web page text, and web page text is handled, obtain web page text word,
Further respectively by the corresponding term vector of web page text word word corresponding with being used to reflect the Ontological concept word of each point of interest to
Amount carries out similarity calculation, obtains the similarity between each web page text word and each Ontological concept word, and remain larger than pre-
If the similarity of threshold value, the similarity between web page text word and each Ontological concept word browse after adding up divided by user
Webpage quantity obtains interest value of the user under each Ontological concept word.
For the embodiment of the present invention, webpage text specifically is being extracted from the corresponding behavioral data of user's browsing webpage behavior
During this, text structure in webpage can use, the web page text in webpage in different zones is extracted, thus quick and precisely
Ground extracts web page text.Under normal conditions, text sections can be saved in webpage, for example, theme type text block, catalogue
Type text block, picture type text block etc. further extract the web page text in webpage in different text blocks.
102, emerging on each Ontological concept word to the user according to the hierarchical relationship between each Ontological concept word
Interesting value is updated, and obtains effective interest value of the user on each Ontological concept word.
Since there are multiple hierarchical relationships between each Ontological concept word, Ontological concept word can be divided into father's notional word
With sub- notional word, there are multiple sub- notional words under each father's notional word, for example, football belongs to ball, football is that ball son is general
Read word, shuttlecock belong to it is ball, shuttlecock be also be ball sub- notional word.
To with the embodiment of the present invention, specifically update user during the interest value on each Ontological concept word, by
There are the differentiations of father's notional word and sub- notional word between each Ontological concept word, by the way that each Ontological concept word is mapped as layer
Grade mode, includes Ontological concept word in each level, with increasing for level, Ontological concept word is in current level
The sub- notional word of Ontological concept word in a upper level, had both been Ontological concept in current level positioned at Ontological concept word in next level
Word in notional word, also in a upper level Ontological concept word sub- notional word, based between each Ontological concept word
Hierarchical relationship determines each sub- notional word for including under father's notional word, the level crossed between every sub- notional word and father's notional word
Difference may illustrate that the degree of correlation between father's notional word is wanted for 1 layer or multilayer, the sub- notional word bigger for level difference
Small, the level being based further between each sub- notional word and father's notional word is poor, is that additional interest value is arranged in each sub- notional word,
The additional interest value has invigoration effect for father's notional word, and then this is added interest value and is added on father's notional word, accordingly
Effective interest value after being updated on each Ontological concept word.
Existing user interest portrait building process in, often will appear point of interest of the user on sub- notional word compared with
Height, and on father's notional word and be not present, the user interest portrait inaccuracy of building is allowed in this way, is led in the embodiment of the present invention
It crosses and user is added to interest value of the user on father's notional word in the interest value on sub- notional word, since sub- notional word and father are general
Word is read with relevance, so that father's notional word equally has corresponding interest value, improves the accuracy of building user interest portrait.
103, effective interest value based on the user on each Ontological concept word, building record the use of each point of interest
Family interest portrait.
Since effective interest value under each Ontological concept word is summarize under the Ontological concept word each sub- notional word attached
Interest value after adding interest value, further effective interest value according to user on each Ontological concept word, being formed in record has
The user interest of each point of interest is drawn a portrait, so that the interest value in user interest portrait under each Ontological concept word is more reasonable.
104, according to the effective interest value for recording each point of interest in user interest portrait, by effective interest value
Ranking is before default value and application relevant to point of interest is recommended to user.
For the embodiment of the present invention, constructs user interest portrait and be equivalent to the process to label to user, specifically exist
Construct user interest portrait during, due to user the virtual value on each Ontological concept word with user browse webpage in
Appearance may be variation, by the way that virtual value of the user on each Ontological concept word is carried out mark to user as dynamic data
Label label to user based on virtual value of the user on each Ontological concept word for accurate user interest portrait
On the basis of, it needs to label to user as static data in conjunction with attribute datas such as name, gender, the ages of user,
Accurately user interest is drawn a portrait for building, and then effective interest value ranking based on each point of interest in user interest portrait is default
Before numerical value and application relevant to point of interest recommends into user, such as user interest portrait record user in sport, fortune
Interest value on dynamic is in the top, can search for application relevant with sport, movement from the background, and will be suitably using recommending extremely use
Family.
The embodiment of the present invention provides a kind of application recommended method based on user interest portrait, each for reflecting by obtaining
The Ontological concept word of a point of interest, and the web page text word that user browses webpage behavior is mapped to and is used to reflect corresponding point of interest
Ontological concept word on, interest value of the user on each Ontological concept word is obtained, according to the layer between each Ontological concept word
Grade relationship is updated interest value of the user on each Ontological concept word, obtains user's having on each Ontological concept word
Interest value, the validity based on user on each Ontological concept word are imitated, building records user's portrait of each interest value, according to
Effective interest value that each point of interest is recorded in user interest portrait, recommends user for related application.With it is straight in the prior art
The user interest portrait constructed on the Ontological concept for being mapped to corresponding point of interest based on the web page text for browsing user is connect to carry out
The mode that user's application is recommended is compared, between the Ontological concept in the user interest portrait of the prior art for characterizing point of interest simultaneously
There is no hierarchical relationship, the present invention fully takes into account the hierarchical relationship between each Ontological concept word concept word, is based on sub- concept
Word is updated the influence degree of father's notional word to interest value of the user on father's notional word, to guarantee that the user of building is emerging
Interest portrait is more reasonable, improves the accuracy that user's application is recommended based on user interest portrait.
The embodiment of the invention provides the application recommended method that another kind is drawn a portrait based on user interest, can construct more rationally
User interest portrait, improve based on user interest portrait recommends user application accuracy, as shown in Fig. 2, the method
Include:
201, the Ontological concept word for reflecting each point of interest is obtained, and user is browsed to the web page text of webpage behavior
Word carries out similarity calculation with the Ontological concept word for reflecting each point of interest, obtains web page text word and each Ontological concept
Similarity between word.
It should be noted that the overdue web page browsing behavior for hitting generation in order to prevent, can browse webpage generating user
When behavior, webpage behavior can be browsed to user and screened, retaining user in the browsing pages residence time is more than preset time
Web page browsing behavior.
For the embodiment of the present invention, specifically can by by user browse webpage this cliction of webpage be used to reflect it is each emerging
The Ontological concept word of interest point is expressed as the corresponding term vector of web page text word and is used to reflect that the ontology of each point of interest to be general
The corresponding term vector of word is read, the Ontological concept for calculating the corresponding term vector of web page text word with being used to reflect each point of interest is passed through
Similarity between the corresponding term vector of word obtains the similarity between web page text word and each Ontological concept word.For example, can
To measure the similarity between word by calculating the included angle cosine value between two term vectors, here without limiting.
202, divided by user after the similarity for being greater than preset threshold between web page text word and each Ontological concept word that adds up
Webpage quantity is browsed, interest value of the user on each Ontological concept word is obtained.
For the embodiment of the present invention, due to being greater than the similar of preset threshold between web page text word and each Ontological concept word
Degree is able to reflect the similarity degree between web page text word and Ontological concept word, and similarity degree is higher, shows user to the ontology
Interest on notional word is higher, and user may jump multiple webpages during browsing webpage, and each webpage can mention
Multiple web page text words are taken, after being greater than the similarity of preset threshold between cumulative web page text word and each Ontological concept word
Webpage quantity is browsed divided by user, obtains interest value of the user on Ontological concept word.
For example, from user browse the web page text word extracted in the corresponding behavioral data of a certain webpage behavior have a, b, c,
D, e, f, g calculate separately the similarity between a, b, c, d, e, f, g and each Ontological concept, remain larger than 0.7 similarity and have
A and sport Ontological concept word are 0.8, b and music Ontological concept word is 0.7, d and sport Ontological concept word is 0.9, f and sport
Ontological concept word is 0.8, is 1 since user browses webpage quantity, calculates interest value of the user under music Ontological concept and is
0.7, the interest value under sport Ontological concept is 0.8+0.9+0.8=2.5.
203, Ontological concept word is divided into father's notional word and sub- notional word with hierarchical relationship, each father's concept in advance
There are multiple sub- notional words under word.
It should be noted that during Ontological concept word is divided into father's notional word and sub- notional word, Ontological concept word
It may be both father's notional word and sub- notional word when towards different Ontological concept words, for example, ball when towards football, football category
In ball, ball is father's notional word of football, and ball when towards sport, ball to belong to sport, it is ball be sport son it is general
Read word.
For the embodiment of the present invention, it is generally the case that there are multiple sub- notional words under each father's notional word, for example, body
Educate as father's notional word there may be it is ball, swimming, running etc. multiple sub- notional words.And sub- notional word is equally possible to have different layers
Multiple father's notional words of grade, for example, shuttlecock is as sub- notional word, there may be multiple father's notional words such as ball, sport, movement.
204, according to the hierarchical relationship between each Ontological concept word, each sub- concept for including under father's notional word is determined
Word.
For the embodiment of the present invention, for Ontological concept word division need to fully take into account towards different Ontological concepts
Word, although the level of father's notional word is higher than sub- notional word, different father's notional words equally may be other Ontological concept words
Sub- notional word, so between father's notional word still also have hierarchical relationship can be determined herein for each father's notional word
Belong to each sub- notional word of father's notional word.
It is understood that here other than the Ontological concept word of lowest hierarchical level, the Ontological concept word of each level
Can be used as father's notional word includes multiple sub- notional words.
205, poor based on the level under father's notional word between each sub- notional word for including and father's notional word, it calculates each
Additional interest value of a sub- notional word on father's notional word.
Wherein, the level difference between each sub- notional word for including under father's notional word and father's notional word is for showing that the son is general
Read accounting situation of the word on father's notional word, that is, as the level difference between fruit notional word and father's notional word is bigger, explanation
The accounting of the sub- notional word on father's notional word is smaller, for example, the ball level difference between movement is 1, and shuttlecock and fortune
Level difference between dynamic is 2, illustrates that shuttlecock is less than the ball accounting in movement moving upper accounting, in certain same level
Level difference of the sub- notional word usually between father's notional word it is identical, for example, shuttlecock, basketball are all ball sub- notional words,
So the level difference between shuttlecock, basketball and movement is all 2.
For the embodiment of the present invention, specifically can according to each sub- notional word including under father's notional word and father's notional word it
Between level it is poor, determine level accounting coefficient of each sub- notional word on father's notional word, which is able to reflect sub general
Word is read to influence degree on father's notional word, calculates separately layer and accounting coefficient and user of each sub- notional word on father's notional word
Product between the interest value on each sub- notional word obtains the additional interest value on each sub- notional word.
For the embodiment of the present invention, above-mentioned level accounting coefficient and level difference inversely, usually can be according to fathers
Level between each sub- notional word for including under notional word and father's notional word is poor, the differential inverse of generation layer, and computation layer is differential
It is reciprocal and preset the product in level between the level parameter of each sub- notional word, determine each sub- notional word in father's concept
Level accounting coefficient on word.
It should be noted that presetting in level the layer and parameter of each sub- notional word commonly used in reflecting the sub- notional word
It, specifically can be according to similar between each sub- notional word and father's notional word in level with the similarity degree between father's notional word
Degree, determines the level parameter of each sub- notional word in level, can also be general by the son in same level of course for calculating is convenient for
It reads word and is set as identical level parameter, here without limiting.
For example, include 5 sub- notional word a1, a2, a3, a4, a5 under father's notional word A, and sub- notional word a1 and a2 with
Level difference between father's notional word A is that the level difference between 1, a3 and father's notional word A is 2, between a4, a5 and father's notional word A
Level difference is 3, and the level parameter where presetting a1, a2 in level is n1, and the level parameter where a3 in level is n2,
Level parameter where a4, a5 in level is n3, it is determined that the level accounting coefficient of sub- notional word a1, a2 on father's notional word A
For 1 × n1, level accounting coefficient of the sub- notional word a3 on father's notional word A is 1/2 × n2, and sub- notional word a4, a5 are general in father
The level accounting coefficient read on word A is 1/3 × n3, further calculates separately interest value and 1 of the user on sub- notional word a1, a2
The product of × n1, user are in the product of interest value and 1/2 × n2 on sub- notional word a3, user on sub- notional word a4, a5
The product of interest value and 1/3 × n3 obtains the additional interest of each sub- notional word a1, a2, a3, a4, a5 on father's notional word A
Value.
206, additional interest value of each sub- notional word on father's notional word is added to user on father's notional word
Interest value, effective interest value after being updated on each Ontological concept word.
For the embodiment of the present invention, usually there are multiple sub- notional words in father's notional word, calculate every sub- notional word
Additional interest value of the sub- notional word on father's notional word, by by the every additional interest value of sub- notional word on father's notional word
It is added to interest value of the user on father's notional word, thus interest value of the user on father's notional word after being updated, for every
A father's notional word is carried out above-mentioned update operation, obtains updated interest value on each Ontological concept word, this is updated
Interest value can more acurrate interest value of the reflection user on each Ontological concept word, be further used as the effective of Ontological concept word
Interest value.
207, effective interest value based on the user on each Ontological concept word, building record the use of each point of interest
Family interest portrait.
Further include the attribute data of user for the embodiment of the present invention, in user interest portrait, can specifically pass through user
The registration information of Website login is available to arrive such as birthday, gender, address, hobby user attribute data, in conjunction with the category of user
Property data and user on each Ontological concept word effective interest value building user interest portrait.
208, according to the effective interest value for recording each point of interest in user interest portrait, by effective interest value
Ranking is before default value and application relevant to point of interest is recommended to user.
As practical application, user interest portrait can do recommender system, according to user on each Ontological concept
Related content is pushed to user by interest value, for example, building user interest portrait shows that interest value of the user in makeups is very high,
The makeups product of skin care item etc is then pushed to user.
Further, the specific implementation as Fig. 1 the method, the embodiment of the invention provides one kind to be based on user interest
Portrait applies recommendation apparatus, as shown in figure 3, described device includes: map unit 31, updating unit 32, construction unit 33, pushes away
Recommend unit 34.
Map unit 31 can be used for obtaining the Ontological concept word for reflecting each point of interest, and user browsed net
The web page text word that page line is is mapped to the Ontological concept word for being used to reflect each point of interest, obtains user at each
Interest value on body notional word；
Updating unit 32 can be used for according to the hierarchical relationship between each Ontological concept word to the user at each
Interest value on body notional word is updated, and obtains effective interest value of the user on each Ontological concept word；
Construction unit 33 can be used for effective interest value based on the user on each Ontological concept word, building note
Record the user interest portrait of each point of interest；
Recommendation unit 34 can be used for according to the effective interest value for recording each point of interest in user interest portrait,
By effective interest value ranking before default value and application relevant to point of interest is recommended to user.
A kind of application recommendation apparatus based on user interest portrait provided by the invention, it is each emerging for reflecting by obtaining
The Ontological concept word of interest point, and the web page text word that user browses webpage behavior is mapped to the sheet for being used to reflect corresponding point of interest
On body notional word, interest value of the user on each Ontological concept word is obtained, is closed according to the level between each Ontological concept word
System is updated interest value of the user on each Ontological concept word, and it is effective emerging on each Ontological concept word to obtain user
Interest value, the validity based on user on each Ontological concept word, building records user's portrait of each interest value, according to user
Effective interest value that each point of interest is recorded in interest portrait, recommends user for related application.With direct base in the prior art
User is carried out in the user interest portrait constructed on the Ontological concept that the web page text for browsing user is mapped to corresponding point of interest
It compares using the mode of recommendation, is not deposited between the Ontological concept in the user interest portrait of the prior art for characterizing point of interest
In hierarchical relationship, the present invention fully takes into account the hierarchical relationship between each Ontological concept word concept word, is based on sub- notional word pair
The influence degree of father's notional word is updated interest value of the user on father's notional word, to guarantee that the user interest of building is drawn
As more reasonable, the accuracy of user's application is recommended in raising based on user interest portrait.
As the further explanation using recommendation apparatus based on user interest portrait shown in Fig. 3, Fig. 4 is according to this
The structural schematic diagram using recommendation apparatus that inventive embodiments another kind is drawn a portrait based on user interest, as shown in figure 4, described device
Further include:
Division unit 35 can be used for existing to the user in the hierarchical relationship according between each Ontological concept word
Interest value on each Ontological concept word is updated, and obtains user before the update interest value on each Ontological concept word,
Ontological concept word is divided into father's notional word with hierarchical relationship and sub- notional word in advance, there are multiple under each father's notional word
Sub- notional word.
Further, the updating unit 32 includes:
Determining module 321 can be used for determining and wrapping under father's notional word according to the hierarchical relationship between each Ontological concept word
Each sub- notional word contained；
First computing module 322 can be used for based on each sub- notional word and father's concept for including under father's notional word
Level between word is poor, calculates additional interest value of each sub- notional word on father's notional word；
Adding module 323 can be used for for additional interest value of each sub- notional word on father's notional word being added to
Interest value of the user on father's notional word, effective interest value after being updated on each Ontological concept word.
Further, first computing module 322 includes；
It determines submodule 3221, can be used for according to each sub- notional word and father's notional word for including under father's notional word
Between level it is poor, determine level accounting coefficient of each sub- notional word on father's notional word；
Computational submodule 3222 can be used for calculating separately level accounting of each sub- notional word on father's notional word
The product of coefficient and user between the interest value on each sub- notional word, it is attached on father's notional word to obtain each sub- notional word
Add interest value.
Further, the determining submodule 3221 specifically can be used for each according to include under father's notional word
Level between sub- notional word and father's notional word is poor, the differential inverse of generation layer；
The determining submodule 3221 specifically can be also used for calculating the level difference inverse and presetting in level respectively
Product between the level parameter of a sub- notional word determines level accounting coefficient of each sub- notional word on father's notional word.
Further, the determining submodule 3221, specifically can be also used for it is described calculate level difference it is reciprocal with
The product in level between the level parameter of each sub- notional word is preset, determines each sub- notional word on father's notional word
Before level accounting coefficient, according to the similarity in level between each sub- notional word and father's notional word, determine in the level
The level parameter of each sub- notional word.
Further, the map unit 31 includes:
Second computing module 311, can be used for by user browse webpage behavior web page text word be used to reflect it is each
The Ontological concept word of point of interest carries out similarity calculation, obtains the similarity between web page text word and each Ontological concept word；
Accumulator module 312 can be used for being greater than preset threshold between cumulative web page text word and each Ontological concept word
Webpage quantity is browsed divided by user after similarity, obtains interest value of the user on each Ontological concept word.
It should be noted that each involved by a kind of application recommendation apparatus based on user interest portrait provided in this embodiment
Other corresponding descriptions of functional unit, can be with reference to the corresponding description in Fig. 1 and Fig. 2, and details are not described herein.
It is deposited thereon based on above-mentioned method as depicted in figs. 1 and 2 correspondingly, the present embodiment additionally provides a kind of storage medium
Computer program is contained, which realizes that the above-mentioned user interest that is based on as shown in Fig. 1 and Fig. 2 is drawn a portrait when being executed by processor
Application recommended method.
Based on this understanding, the technical solution of the application can be embodied in the form of software products, which produces
Product can store in a non-volatile memory medium (can be CD-ROM, USB flash disk, mobile hard disk etc.), including some instructions
With so that computer equipment (can be personal computer, server or the network equipment an etc.) execution the application is each
Method described in implement scene.
Based on above-mentioned method as shown in Figure 1 and Figure 2 and Fig. 3, virtual bench embodiment shown in Fig. 4, in order to realize
Above-mentioned purpose, the embodiment of the present application also provides a kind of computer equipments, are specifically as follows personal computer, server, network
Equipment etc., the entity device include storage medium and processor；Storage medium, for storing computer program；Processor is used for
Computer program is executed to realize the above-mentioned application recommended method based on user interest portrait as depicted in figs. 1 and 2.
Optionally, which can also include user interface, network interface, camera, radio frequency (Radio
Frequency, RF) circuit, sensor, voicefrequency circuit, WI-FI module etc..User interface may include display screen
(Display), input unit such as keyboard (Keyboard) etc., optional user interface can also connect including USB interface, card reader
Mouthful etc..Network interface optionally may include standard wireline interface and wireless interface (such as blue tooth interface, WI-FI interface).
It will be understood by those skilled in the art that it is provided in this embodiment based on user interest portrait using recommendation apparatus
Entity device structure does not constitute the restriction to the entity device, may include more or fewer components, or combination is certain
Component or different component layouts.
It can also include operating system, network communication module in storage medium.Operating system is that the above-mentioned computer of management is set
The program of standby hardware and software resource, supports the operation of message handling program and other softwares and/or program.Network communication mould
Block leads to for realizing the communication between each component in storage medium inside, and between other hardware and softwares in the entity device
Letter.
Through the above description of the embodiments, those skilled in the art can be understood that the application can borrow
It helps software that the mode of necessary general hardware platform is added to realize, hardware realization can also be passed through.Pass through the skill of application the application
The web page text of user's browsing is mapped to structure on the Ontological concept of corresponding point of interest to being directly based upon in the prior art by art scheme
The user interest portrait built carries out the mode that user's application is recommended and compares, emerging for characterizing in the user interest portrait of the prior art
Between the Ontological concept of interest point and hierarchical relationship is not present, the present invention fully takes into account between each Ontological concept word concept word
Hierarchical relationship is updated interest value of the user on father's notional word based on influence degree of the sub- notional word to father's notional word,
User interest portrait to guarantee building is more reasonable, improves the accuracy that user's application is recommended based on user interest portrait.
It will be appreciated by those skilled in the art that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, module in attached drawing or
Process is not necessarily implemented necessary to the application.It will be appreciated by those skilled in the art that the mould in device in implement scene
Block can according to implement scene describe be distributed in the device of implement scene, can also carry out corresponding change be located at be different from
In one or more devices of this implement scene.The module of above-mentioned implement scene can be merged into a module, can also be into one
Step splits into multiple submodule.
Above-mentioned the application serial number is for illustration only, does not represent the superiority and inferiority of implement scene.Disclosed above is only the application
Several specific implementation scenes, still, the application is not limited to this, and the changes that any person skilled in the art can think of is all
The protection scope of the application should be fallen into.
Claims (10)
1. a kind of application recommended method based on user interest portrait, which is characterized in that the described method includes:
The Ontological concept word for reflecting each point of interest is obtained, and the web page text word that user browses webpage behavior is mapped to
The Ontological concept word for being used to reflect each point of interest, obtains interest value of the user on each Ontological concept word；
Interest value of the user on each Ontological concept word is carried out according to the hierarchical relationship between each Ontological concept word
It updates, obtains effective interest value of the user on each Ontological concept word；
Effective interest value based on the user on each Ontological concept word, the user interest that building records each point of interest are drawn
Picture；
According to the effective interest value for recording each point of interest in user interest portrait, by effective interest value ranking pre-
If before numerical value and application relevant to point of interest is recommended to user.
2. the method according to claim 1, wherein being closed in the level according between each Ontological concept word
System is updated interest value of the user on each Ontological concept word, obtains user on each Ontological concept word more
Before new interest value, the method also includes:
Ontological concept word is divided into father's notional word and sub- notional word with hierarchical relationship in advance, is existed under each father's notional word
Multiple sub- notional words.
3. according to the method described in claim 2, it is characterized in that, the hierarchical relationship according between each Ontological concept word
Interest value of the user on each Ontological concept word is updated, it is effective on each Ontological concept word to obtain user
Interest value includes:
According to the hierarchical relationship between each Ontological concept word, each sub- notional word for including under father's notional word is determined；
It is poor based on the level under father's notional word between each sub- notional word for including and father's notional word, calculate each sub- concept
Additional interest value of the word on father's notional word；
Additional interest value of each sub- notional word on father's notional word is added to interest value of the user on father's notional word,
Effective interest value after being updated on each Ontological concept word.
4. according to the method described in claim 3, it is characterized in that, described general based on each height for including under father's notional word
The level read between word and father's notional word is poor, calculates additional interest value of each sub- notional word on father's notional word and includes:
It is poor according to the level between each sub- notional word for including under father's notional word and father's notional word, determine each sub- concept
Level accounting coefficient of the word on father's notional word；
Calculate separately each sub- notional word on father's notional word level accounting coefficient and user on each sub- notional word
Interest value between product, obtain additional interest value of each sub- notional word on father's notional word.
5. according to the method described in claim 4, it is characterized in that, described general according to each height for including under father's notional word
The level read between word and father's notional word is poor, determines that level accounting coefficient of each sub- notional word on father's notional word includes:
Poor according to the level between each sub- notional word for including under father's notional word and father's notional word, generation layer is differential
Number；
It calculates the level difference inverse and presets the product in level between the level parameter of each sub- notional word, determine each
Level accounting coefficient of a sub- notional word on father's notional word.
6. according to the method described in claim 5, it is characterized in that, in the calculating level difference inverse and presetting layer
Product in grade between the level parameter of each sub- notional word determines level accounting system of each sub- notional word on father's notional word
Before number, the method also includes:
According to the similarity in level between each sub- notional word and father's notional word, each sub- notional word is determined in the level
Level parameter.
7. according to method described in right 1, which is characterized in that the web page text word that user is browsed webpage behavior is mapped to
The Ontological concept word for being used to reflect each point of interest, obtaining interest value of the user on each Ontological concept word includes:
The web page text word that user browses webpage behavior is subjected to similarity with the Ontological concept word for being used to reflect each point of interest
It calculates, obtains the similarity between web page text word and each Ontological concept word；
It is greater than after the similarity of preset threshold between cumulative web page text word and each Ontological concept word and browses webpage divided by user
Quantity obtains interest value of the user on each Ontological concept word.
8. a kind of apply recommendation apparatus based on user interest portrait, which is characterized in that described device includes:
User for obtaining the Ontological concept word for reflecting each point of interest, and is browsed the net of webpage behavior by map unit
Page text word is mapped to the Ontological concept word for being used to reflect each point of interest, obtains user on each Ontological concept word
Interest value；
Updating unit, for according to the hierarchical relationship between each Ontological concept word to the user in each Ontological concept word
Interest value be updated, obtain effective interest value of the user on each Ontological concept word；
Construction unit, for effective interest value based on the user on each Ontological concept word, building records each interest
The user interest portrait of point；
Recommendation unit will be described effective for recording effective interest value of each point of interest in drawing a portrait according to the user interest
Interest value ranking is before default value and application relevant to point of interest is recommended to user.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 7 the method when executing the computer program.
10. a kind of computer storage medium, is stored thereon with computer program, which is characterized in that the computer program is located
The step of reason device realizes method described in any one of claims 1 to 7 when executing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910319420.9A CN110209908A (en) | 2019-04-19 | 2019-04-19 | Application recommended method, device, computer equipment and computer storage medium based on user interest portrait |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910319420.9A CN110209908A (en) | 2019-04-19 | 2019-04-19 | Application recommended method, device, computer equipment and computer storage medium based on user interest portrait |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110209908A true CN110209908A (en) | 2019-09-06 |
Family
ID=67786046
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910319420.9A Pending CN110209908A (en) | 2019-04-19 | 2019-04-19 | Application recommended method, device, computer equipment and computer storage medium based on user interest portrait |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110209908A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111310074A (en) * | 2020-02-13 | 2020-06-19 | 北京百度网讯科技有限公司 | Method, apparatus, electronic device, and computer-readable medium for tag optimization of points of interest |
| CN111708901A (en) * | 2020-06-19 | 2020-09-25 | 腾讯科技（深圳）有限公司 | Multimedia resource recommendation method and device, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140122456A1 (en) * | 2012-03-30 | 2014-05-01 | Percolate Industries, Inc. | Interactive computing recommendation facility with learning based on user feedback and interaction |
| US20170212949A1 (en) * | 2016-01-21 | 2017-07-27 | Adobe Systems Incorporated | Auditing and Augmenting User-Generated Tags for Digital Content |
| CN108288229A (en) * | 2018-03-02 | 2018-07-17 | 北京邮电大学 | A kind of user's portrait construction method |
| CN108920521A (en) * | 2018-06-04 | 2018-11-30 | 上海财经大学 | User's portrait-item recommendation system and method based on pseudo- ontology |
-
2019
- 2019-04-19 CN CN201910319420.9A patent/CN110209908A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140122456A1 (en) * | 2012-03-30 | 2014-05-01 | Percolate Industries, Inc. | Interactive computing recommendation facility with learning based on user feedback and interaction |
| US20170212949A1 (en) * | 2016-01-21 | 2017-07-27 | Adobe Systems Incorporated | Auditing and Augmenting User-Generated Tags for Digital Content |
| CN108288229A (en) * | 2018-03-02 | 2018-07-17 | 北京邮电大学 | A kind of user's portrait construction method |
| CN108920521A (en) * | 2018-06-04 | 2018-11-30 | 上海财经大学 | User's portrait-item recommendation system and method based on pseudo- ontology |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111310074A (en) * | 2020-02-13 | 2020-06-19 | 北京百度网讯科技有限公司 | Method, apparatus, electronic device, and computer-readable medium for tag optimization of points of interest |
| CN111310074B (en) * | 2020-02-13 | 2023-08-18 | 北京百度网讯科技有限公司 | Method and device for optimizing labels of interest points, electronic equipment and computer readable medium |
| US12020467B2 (en) | 2020-02-13 | 2024-06-25 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for optimizing tag of point of interest, electronic device and computer readable medium |
| CN111708901A (en) * | 2020-06-19 | 2020-09-25 | 腾讯科技（深圳）有限公司 | Multimedia resource recommendation method and device, electronic equipment and storage medium |
| CN111708901B (en) * | 2020-06-19 | 2023-10-13 | 腾讯科技（深圳）有限公司 | Multimedia resource recommendation method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200117675A1 (en) | Obtaining of Recommendation Information | |
| US20230047611A1 (en) | Method for displaying search contents and electronic device | |
| CN101436186B (en) | Method and system for providing related searches | |
| US7555478B2 (en) | Search results presented as visually illustrative concepts | |
| Feng et al. | Mining user-contributed photos for personalized product recommendation | |
| CN112313697A (en) | System and method for generating interpretable description-based recommendations describing angle augmentation | |
| CN110413877A (en) | A kind of resource recommendation method, device and electronic equipment | |
| US20100274753A1 (en) | Methods for filtering data and filling in missing data using nonlinear inference | |
| US20180189405A1 (en) | Method and device for displaying recommended contents | |
| CN110287307B (en) | A kind of search result sorting method, device and server | |
| US20060155751A1 (en) | System and method for document analysis, processing and information extraction | |
| US20140122475A1 (en) | Search result ranking method and system | |
| CN105718184A (en) | Data processing method and apparatus | |
| CN105931082B (en) | Commodity category keyword extraction method and device | |
| US9330071B1 (en) | Tag merging | |
| CN104598556A (en) | Search method and search device | |
| CN105589852B (en) | A kind of method and apparatus of information recommendation | |
| CN112488781A (en) | Search recommendation method and device, electronic equipment and readable storage medium | |
| CN106293119A (en) | A kind of method and apparatus carrying out information recommendation in input method | |
| CN106570031A (en) | Service object recommending method and device | |
| CN104123321B (en) | A kind of determining method and device for recommending picture | |
| CN109299378B (en) | Search result display method and device, terminal and storage medium | |
| WO2025060841A1 (en) | Resource screening method and apparatus, computer device and storage medium | |
| CN110209908A (en) | Application recommended method, device, computer equipment and computer storage medium based on user interest portrait | |
| KR20090120843A (en) | Multi-concept network creation system and method based on user web usage information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190906 |