JP3298288B2 - Document input device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document input device capable of efficiently inputting Japanese by preferentially presenting a candidate selected by a user in Japanese kana-kanji conversion. Things.

[0002]

2. Description of the Related Art When a Japanese document is created, a method of inputting a kana and converting it into a kanji has conventionally been used. Since there are generally a plurality of kanji corresponding to the kana, the user has to repeat the re-conversion until the desired kanji is displayed. In order to reduce the number of re-conversion operations, the most recently used kanji is recorded in a storage area such as a RAM, and when the same reading is input, the kanji recorded in the storage area is preferentially displayed. Is used.

[0003] In this method, there is a problem in what criteria the word selected by the user is registered in the storage area, and how the word is updated when the same word exists in the storage area. Also, in general, the area for recording used kanji is limited, so to register a new word, it is necessary to delete the previously registered word from the storage area,
It also matters which words to delete.

Conventionally, in response to such a demand,
A system as shown in FIG. 20 has been proposed. FIG. 20 is an explanatory diagram for explaining the operations of the word dictionary and the priority information in the kana-kanji conversion processing system disclosed in Japanese Patent Application Laid-Open No. 61-862.

In FIG. 20A, reference numeral 203 denotes a ROM which stores a plurality of word readings and word groups composed of kanji notation words having the readings. As shown in FIG. The words in kanji notation include "climate", "calling port", "contribution", "stomata", etc.
These make up one word group. Reference numeral 204 denotes a RAM area allocated to store priority information, in which a plurality of priority information groups each including a reading of a word and priority information of a word in kanji notation having the reading are stored. The numbers such as 3 and 2 in the priority information group are the kanji notation sequence numbers stored in the word group reading “Kikou”,
The number is "Contribution" stored in the third word group,
The second priority indicates that "port of call" is stored second in the word group. Since only 3 and 2 are registered in the priority information group of the RAM area 204, the subsequent priorities are “climate” and “porosity” in the order stored in the word group of the ROM 203. So in this example, the priorities are "Contribution", "Port of Call",
"Climate" and "porosity".

In this state, when the fourth "porosity" in the word group is determined as a kanji character string for "Kiko", the priority information group "Kiko 4, 3, 2" as shown in FIG. Is written at the beginning of the RAM area 204, and the priority order is “porosity”, “contribution”, “port”,
"Climate". At this time, the priority information group already stored in the RAM area 204 is shifted backward,
If not, the last one is kicked out.

[0007]

Since the conventional document input device is configured as described above, there are the following problems.

First, registration in the priority word dictionary is performed in correspondence with kana reading and kana kanji mixture (hereinafter referred to as kanji mixture) notation for each word, and does not consider the structure of phrases. Was. In the conventional registration in the preferred word dictionary, since the unit of reading is an independent word, for example, the sentence "Kishakushasha" is converted to "Returned by train", and the kanji for the reading "Kisha" is prioritized When registering in the word dictionary, "homecoming" registered later has higher priority, so that the first "by train" cannot be the first candidate. This is because the registration in the preferred word dictionary is not performed for each phrase “Kishade” or “Kishashita”.

Second, deletion from the priority word dictionary, which is performed to make effective use of the storage area of the priority word dictionary, is performed from old words registered in the priority word dictionary, and important words and general words are deleted. We did not distinguish it from words. Since the preferred word dictionary area for recording used kanji is limited, in order to register a new word, it may be necessary to delete a previously registered word from the storage area. For example, with respect to "kisha", different notations such as "car" and "return" often appear in the same sentence in many cases, and there is a high probability that a word registered later and given priority will not be the correct answer. But,
Regarding the personal name “Takahiro”, there are few cases where different notations such as “Takahiro”, “Takahiro”, and “Takahiro” appear in the same sentence, and the notation selected by the user is likely to appear again in the same sentence. Therefore, when deleting from the priority word dictionary, it is better to treat personal names and the like that have a high probability of reappearing as important words, and delete “train” and “return to home” that have a low probability of reappearing first. Words were deleted without distinguishing them from ordinary words.

A candidate having a large amount of operation to obtain a target notation, such as pressing the reconversion key many times, has to repeat many operations again when it is forgotten. It can be said that it is better to give importance to the word than the candidate obtained, but conventionally, the operation amount has not been considered.

Third, the maximum number of homophone candidates to be recorded in the priority word dictionary for one reading is the same for all readings. Since readings with a large number of homophone candidates are likely to be re-converted, it is preferable to memorize a large number of homophones rather than readings with a small number of candidates. In some cases, it may be better not to delete words that require a large amount of operation from the priority word dictionary due to homophones that can be obtained with a small amount of operation. However, conventionally, the number of homophone candidates has not been taken into account when recording or deleting in the priority word dictionary.

Fourth, recording of the most recently used word in the preferred word dictionary does not take into account the connection between words. For example, in the context of "mysterious phenomena", "mystery" takes precedence over other homonyms in the context of being adjacent to the word "phenomenon". For, candidates that are frequently used, such as "kai" and "solution" that are not "mysteries", are likely to be selected.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems. In kana-kanji conversion, words that are used by a user are appropriately displayed, and words used by the user are recorded. To obtain a document input device that effectively uses a storage area to be used.

[0014]

A document input device according to the present invention.
The input means for inputting Japanese sentences in kana, the basic dictionary storing the kanji notation corresponding to kana and the attribute information of this notation, and the basic character string of the kana character string input by the input means described above. A basic dictionary search unit for searching a dictionary, a phrase generation unit for separating an input kana character string into Japanese phrases from information retrieved by the basic dictionary search unit to generate a kanji mixed notation, and displaying candidates of a kanji mixed notation A display means, a priority word dictionary in which independent words of the candidate selected by the input means are recorded from the candidates displayed on the display means, and the priority word dictionary from the reading of the independent words in the phrase divided by the phrase generation means. In a document input device provided with a preferred word dictionary search means for searching a word dictionary, a candidate candidate selected by the input means is recorded in the preferred word dictionary. Determining the severity by the part of speech of the candidate, provided priority determining means for changing the reference time period to be recorded in the priority dictionary in accordance with their importance, the preferred term dictionary search means, by the phrase generating means The preferred word dictionary is searched from the reading of the separated phrases, and the latest use candidate based on the importance is prioritized.

[0015]

Also, the document input device according to the present invention has its advantages.
Whether the previous determination means is registered in the basic dictionary
The priority level is determined from the number of homophones for the attribute words, and the attribute priority level determination means changes the recording period for recording in the priority word dictionary according to the determined importance level.

Further, the document input device according to the present invention has the above advantage.
The priority determining means determines the importance based on the number of homophones for the kana registered in the basic dictionary, and changes the number of homophones recorded in the priority word dictionary according to the determined importance, thereby changing the priority. Period recorded in the word dictionary
This is constituted by a basic registration homonym number determining means for changing the reference between them.

Further, in the document input device according to the present invention, the priority determination means may include an operation amount totalization means for measuring a total of operation amounts input until a target kanji mixture candidate is selected;
The importance is determined from the sum of the operation amounts, the determined importance and the selected candidate are recorded in the priority word dictionary, and a recording period to be recorded in the priority word dictionary is changed according to the recorded importance. And an operation amount priority determining means.

Further, in the document input device according to the present invention, the priority determination means may include an operation amount totalization means for measuring a total of operation amounts input until a target kanji mixture candidate is selected;
The importance is determined from the sum of the operation amounts, the determined importance and the selected candidate are recorded in the priority word dictionary, and the number of homophone words to be recorded in the priority word dictionary is changed according to the recorded importance. By doing so, the
And the operation amount registration homonym number determining means for changing the reference of the period of the operation.

[0020]

[0021]

In the document input device according to the present invention, a kana character string is input by the input means, and the basic dictionary search means searches the basic dictionary for a partial character string of the input kana character string. On the basis of the result, the phrase generation means separates the input kana character string into Japanese phrases and generates a kanji-mixed notation, and the priority word dictionary search means reads the phrase separated by the phrase generation means from the reading of the priority word dictionary. And retrieves the latest use candidate based on the importance, and when the desired notation is selected by the input means from the display means which displays the extracted candidate and the kanji mixed notation generated by the phrase generation means. , The priority determining means selects the candidate item selected by the input means.
The importance is determined by the lyrics , and the selected candidate is recorded in the preferred word dictionary according to the determined importance,
Further, the priority word dictionary is updated according to the importance, and words having high importance are left in the priority word dictionary.

[0022]

Further, the attribute priority determining means of the document input device according to the present invention determines the importance from the number of homophones for the kana registered in the basic dictionary, and according to the determined importance. The recording period to be recorded in the priority word dictionary is controlled so that words of high importance are left in the priority word dictionary for a long time.

Further , the basic registered homophone number determining means of the document input device according to the present invention determines the importance from the number of homophones for the kana registered in the basic dictionary, and determines the importance based on the determined importance. Therefore, the number of homophones recorded in the priority word dictionary is changed, and many words with high importance are left in the priority word dictionary.

In the document input device according to another aspect of the present invention, the operation amount totaling means measures a total of operation amounts input until a target kanji mixture candidate is selected, and determines the operation amount priority. The means determines the importance from the total of the operation amounts, records the determined importance and the selected candidate in the priority word dictionary, and records the priority in the priority word dictionary according to the recorded importance. The recording period is controlled so that words with high importance are left in the priority word dictionary for a long time.

In the document input device according to another aspect of the present invention, the operation amount totaling means measures the total of the operation amounts input until the target kanji mixture candidate is selected, and obtains the operation amount registration homophone number. The determining means determines the importance from the total operation amount, records the determined importance and the selected candidate in the priority word dictionary, and records the priority in the priority word dictionary according to the recorded importance. The number of homonyms is changed, and many words with high importance are left in the priority word dictionary.

[0027]

[0028]

【Example】

Embodiment 1 FIG. FIG. 1 is an overall configuration diagram of a first embodiment of a document input device according to the present invention. In FIG. 1, reference numeral 1 denotes input means such as a keyboard provided with a kana key for inputting a kana character string of a Japanese sentence, a conversion key, an enter key, and the like. 3 denotes kanji notation corresponding to kana and attribute information on words such as part of speech. The stored basic dictionary 4 is a basic dictionary search means for searching the basic dictionary 3 by using a partial character string of the kana character string input by the input means 1 as a key, and 5 is information searched by the basic dictionary search means 4. A phrase generation unit that divides the input kana character string into Japanese phrases and generates a kanji-mixed notation, 6 is a priority word dictionary that records candidates selected by the user, and 7 is a kanji-mixed notation selected by the user. Extraction of information and information on adjunct words, and phrase information extraction for recording the extracted independent words, adjunct words and phrases, and recording the extracted independent words and advocacy words in the priority word dictionary 6 Stage, 8 priority dictionary search means for searching the priority dictionary 6 readings clauses separated by clause generating means 5, 9 basic dictionary 3
Display means such as a display for displaying kanji-mixed notation candidates created from the result of searching the keyword and the preferred word dictionary 6. A control unit 2 includes a basic dictionary search unit 4, a phrase generation unit 5, a phrase information extraction unit 7, and a priority word dictionary search unit 8, and performs processing described below according to a control program written in a ROM or the like. .

FIG. 2 is an explanatory diagram of the preferred word dictionary 6 in this embodiment. The preferred word dictionary 6 stores the notations of the independent part and the attached part extracted from the notation mixed with the reading and the kanji as shown in FIG. FIG. 3 is a flowchart illustrating the operation of the present embodiment. As an example, a case will be described with reference to the flowchart of FIG. 3 in which a sentence "returned to the train" is obtained in response to an input of "successfully done".
It is assumed that no information is registered in the priority word dictionary 6 as an initial state. In addition, in the basic dictionary 3, for the reading "kisha", "reporter" is the first candidate, "train"
Is registered as a second candidate, and "return" is registered as a third candidate.

First, the input means 1 is used to input a kana of "you are happy" (step 31). Next, the basic dictionary 3 is searched by the basic dictionary search means 4 using the partial character string of the input kana character string as a key (step 32). From the search result, the phrase generation unit 5 divides the phrase into two phrases, "Kishade" and "Kishashita" (step 33). The phrase generation is, for example, n
This is performed by a method such as the longest phrase matching method. As a result, "at the reporter", "at the train", "at home", "reporter", "trained"
A phrase such as "I've returned" is generated. Next, the priority word dictionary 6 is searched by the priority word dictionary search means 8 (step 34). Since no information is currently registered in the priority word dictionary 6, "in a reporter" and "reporter" corresponding to the first candidate "reporter" in the basic dictionary 3 are displayed on the display means 9 (step 35). ).

Since the displayed candidate is not the one desired by the user, the user presses the re-conversion key by the input means 1 to instruct the display of the next candidate (step 36). The user responds to “Kishade” until “Train” appears.
Re-conversion is performed for "Kishashita" until "Returned to Home" appears, and when the desired character string appears, the enter key is pressed.

When the enter key is depressed, the phrase information extracting means 7 extracts information on independent words and information on attached words for both the phrases "by train" and "returned to work". Since both the extracted "in train" and "returned to home" are phrases with attached words (step 37), the phrase information extraction means 7 registers both of the phrases with attached words in the preferred word dictionary 6 (step 37). Step 38) and registration of a clause without an attached word (step 39) are performed.

First, registration in the priority word dictionary 6 is "by train"
The phrase “with train” with attached words is registered (step 38), and the phrase “with train” without attached words is registered (step 39).
(A). Next, for "returned to home", registration of the phrase "returned to home" with attached words (step 38),
Registration of the phrase "return to home" without an attached word is performed (step 39). For the reading "kisha", "train" has already been registered, but to prioritize the new notation "return", "train" is deleted and "return" is registered. As a result, the priority word dictionary 6 is as shown in FIG.

In this embodiment, the old registered word "train" is deleted from the priority word dictionary 6, but "train" may be left as the second candidate in the priority word dictionary 6.

Next, in this state, "I was a young lady."
As a result, the phrase generation means 5 generates a phrase corresponding to "Kishade" and a phrase corresponding to "Kishashita". 8 retrieves "by train" and "returned from" from the priority word dictionary 6, so that the desired character string "returned by train" is displayed first.

When a character string with a different auxiliary word, such as "Kishada," is entered, "Go home" corresponding to the reading "Kisha," which is a partial character string of "Kishada," has priority. A search is made from the word dictionary 6, and "Returned to home" is displayed.

When the notation "train" is displayed for the reading "kisha" without an attached word, the phrase is determined to have no attached word in step 37, and the phrase with the attached word in step 38 is displayed. Is not registered, and only “train” corresponding to “kisha” is registered in the priority word dictionary 6 in step 39.

In this embodiment, an example is shown in which the kanji notation is stored in the independent word notation of the priority word dictionary 6 as shown in FIG. 2, but instead of storing the kanji notation, for example, the kanji in the basic dictionary 3 is stored. The address of the notation may be stored, and an address that can specify the kanji notation stored in the basic dictionary 3 may be stored.

Embodiment 2 FIG. FIG. 4 is an overall configuration diagram of a second embodiment of a document input device according to the present invention. Components similar to those of the first embodiment shown in FIG.
In FIG. 4, reference numeral 6a denotes a preferred word dictionary that records candidates selected by the user, 8a denotes a preferred word dictionary search unit that searches the preferred word dictionary 6a from reading of phrases separated by the phrase generation unit 5, and 10 denotes a user-selected dictionary. The importance of the kanji-mixed notation determined is determined from the attribute information on words such as parts of speech registered in the basic dictionary 3, and the recording period in which the candidate selected by the user is left in the priority word dictionary 6a according to the determined importance is determined. This is attribute priority determining means to be changed. 2a is a basic dictionary search means 4, a phrase generation means 5, a priority word dictionary search means 8a
And a control unit configured to perform processing to be described later according to a control program written in a ROM or the like.

FIG. 5 shows a preferred word dictionary 6a in this embodiment.
FIG. The priority word dictionary 6a stores, as shown in FIG. 5, reading of phrases, notation, rank indicating importance, and count indicating oldness of registration, as shown in FIG. In the rank, A represents an important word and B represents a general word. It is assumed that a common noun, which is an attribute registered in the basic dictionary 3, belongs to the general language rank B and a personal name belongs to the important word rank A. The count indicates the age of the words registered in the priority word dictionary 6a, and a larger value indicates that the registration in the priority word dictionary 6a is older.
For example, "Kisha" for "train" (attribute is a common noun), "Takahiro" for "Takahiro" (attribute is a person's name),
For "kojin", "deceased" (attribute is a common noun) is continuously selected. It is assumed that no information is registered in the priority word dictionary 6a as an initial state.
It is assumed that 00 words can be registered.

FIG. 6 is a flowchart showing the operation of this embodiment. The operation of this embodiment will be described with reference to the flowchart of FIG. First, a description will be given of a case where “train” is selected by reading “kisha”. As in the first embodiment, input of kana (step 61), retrieval of the basic dictionary 3 (step 62), generation of a phrase (step 63), retrieval of the priority word dictionary 6a (step 64), and display of candidates (step 65) ) And re-conversion (step 66). Since “train” has been selected in the above processing, “train” is registered in the priority word dictionary 6a. Since no information is currently registered in the priority word dictionary 6a (step 67), the word is not deleted (step 68), and the count of the words in the priority word dictionary 6a is incremented (step 69). Absent. Since "car" is a common noun, it is determined by the attribute priority determining means 10 as a general word of rank B, and the rank is B and the count is registered as an initial value of 0 in the priority word dictionary 6a (step 70). As a result, the state of the priority word dictionary 6a is as shown in FIG.

Next, the case where "Takahiro" is selected by reading "Takahiro" will be described. Since there is still a storage area to be registered in the priority word dictionary 6a, the same processing as the above-mentioned reading of "Kisha" is performed up to step 67. Next, at step 69, the count of the words in the priority word dictionary 6a is incremented. In this case, since the word in the priority word dictionary 6a is the general word "train", the count of "train" is increased by two. "Takahiro" to be newly registered has the attribute "person name"
Therefore, the keyword is registered as an important word of rank A in the priority word dictionary 6a with an initial value of 0 (step 70). As a result, the state of the priority word dictionary 6a is as shown in FIG.

Further, if "deceased" is selected in the reading of "kojin", the same processing as in the above-mentioned reading of "kisha" is performed up to step 67, and in step 69, the common word "train" is added. The count is increased by two, and the count is increased by one for the key word "Takahiro". "deceased"
Is a general word, it is a general word of rank B, and the count is registered in the priority word dictionary 6a with an initial value of 0 (step 7).
0). As a result, the state of the priority word dictionary 6a is as shown in FIG.

As described above, when a new word is registered, the count of general words already registered in the priority word dictionary 6a is 2
As the number of important words is increased by one, the number of registered words in the priority word dictionary 6a eventually exceeds 300, and there is no storage area to be registered. If you run out of storage space,
In step 68, the word having the largest count value is deleted, so that the important word remains in the priority word dictionary 6a for a longer period than the general word. Further, the deletion of the word having the largest count value is a process necessary for effectively using the storage area of the priority word dictionary 6a. According to the present embodiment, a general word having a low importance is placed before an important word. Can be deleted.

In the present embodiment, the important word is left in the priority word dictionary 6a for a longer period than the general word, depending on the value of the count. Even if it is divided, the important words can be left in the priority word dictionary 6a for a long time. As an example, it is assumed that the number of general words is about 10 times greater than the number of important words in the basic dictionary 3. Further, as shown in the explanatory diagram of the priority word dictionary 6a shown in FIG. 7, if the number of words that can be registered is 300 words, of which 100 words are assigned to important words and 200 words are registered to general words, the basic dictionary 3 Is greater than the ratio of the number of important words to the common words existing in the preferred word dictionary 6a, so that the important words are registered in the preferred word dictionary 6a longer. You can keep.

In this embodiment, personal names are used as important words. However, place names and other proper nouns may be used as important words.
Important words may be determined based on other attributes such as technical terms. Further, in the present embodiment, the words are classified into the important words and the general words. However, the words may be classified into three or more classifications according to the importance.
Further, in this embodiment, 300 words can be registered in the priority word dictionary 6a. However, the maximum number may be more than 300 words, and may be determined by the size of the storage area of the priority word dictionary 6a.

Although the preferred word dictionary 6a according to the present embodiment stores kanji notation as shown in FIG. 5, the kanji notation is stored in the basic dictionary 3 instead of the kanji notation. Or an address that can specify the kanji notation stored in the basic dictionary 3. Although the priority word dictionary 6a according to the present embodiment stores the rank indicating the importance as shown in FIG. 5, the rank does not have to be stored in the priority word dictionary 6a. Since the rank indicating the degree of importance is determined by an attribute relating to words such as part of speech stored in the basic dictionary 3, when the count of the priority word dictionary 6a is updated in step 69 in FIG. The degree may be obtained.

Embodiment 3 FIG. FIG. 8 is an overall configuration diagram of a third embodiment of the document input device according to the present invention. Components similar to those of the first embodiment shown in FIG.
In FIG. 8, reference numeral 6b denotes a preferred word dictionary that records candidates selected by the user, 8b denotes a preferred word dictionary search unit that searches the preferred word dictionary 6b from reading of phrases separated by the phrase generation unit 5, and 11 denotes a user-selected dictionary. Basic registration homonym number determining means for determining the importance of the kanji-mixed notation based on the number of homonyms for kana registered in the basic dictionary 3, and changing the number of homonyms registered in the priority word dictionary 6b according to the determined importance It is. A control unit 2b includes a basic dictionary search unit 4, a phrase generation unit 5, a priority word dictionary search unit 8b, and a basic registration homonym number determination unit 11, and performs a process described below in accordance with a control program written in a ROM or the like. It is.

FIG. 9 shows a preferred word dictionary 6b in this embodiment.
FIG. The priority word dictionary 6b stores two phrases, notation 1 and notation 2, as shown in FIG. In response to the input of the character string written in the reading, the word registered in the notation 1 is displayed on the display means 9 as the first word.
The word displayed as a candidate and registered in the notation 2 is displayed on the display means 9 as a second candidate. As an example, in the basic dictionary 3, for the reading "Ko", "Gong" and "Item"
It is assumed that there are ten homonyms such as "A". In addition, it is assumed that there are two homophone words of “deceased” and “individual” for the reading “kojin”. Note that readings with a large number of homonyms are likely to be re-converted, and it is preferable to remember many homonyms in the priority word dictionary 6b rather than readings with a small number of homonyms. Basic dictionary 3
If seven or more words are registered, it is determined that the degree of importance is high and there are many homonyms, and it is assumed that up to two homonyms can be registered for one reading in the priority word dictionary 6b. It is also assumed that no information is registered in the priority word dictionary 6b as an initial state.

FIG. 10 is a flowchart showing the operation of this embodiment. The operation of this embodiment will be described with reference to the flowchart of FIG. First, a case in which “Ko” is selected by reading “Ko” will be described. As in the first embodiment, input of kana (step 101), retrieval of the basic dictionary 3 (step 102), generation of a phrase (step 10)
3), retrieval of the priority word dictionary 6b (Step 104), display of candidates (Step 105), reconversion (Step 106)
Is performed. Here, the number of homonyms of “ko” is ten and seven or more, so that the basic registration homonym number determining means 11
It is determined that the number of homonyms is large (step 10).
7), “K” is registered in the priority word dictionary 6b (step 108). For the pronunciation of “Ko”, up to two words, notation 1 and notation 2, can be registered in the preferred word dictionary 6b, but at this time, no word is registered for the reading of “Ko”, As shown in FIG. 9A, it is registered in notation 1.

Next, when "item" is selected for "this", the same processing as described above is performed up to step 107. Next, in step 108, "item" is registered. The notation for the reading "Ko" is the priority word dictionary 6b.
9, a new “item” is registered in the notation 1 by the basic registration homonym number determining means 11 as shown in FIG. Further, when "Ko" is selected for "Kou", "Ko" is registered in the notation 1 in step 108, "Item" is shifted to the notation 2, and "Gong" is deleted from the preferred word dictionary 6b. 9 (c).

Next, the case of registering a word corresponding to the reading "kojin" will be described. When registering "deceased", the registration is performed in step 109 because the number of homonyms for the reading "kojin" is two and less than seven in step 107. At this time, since no word has been registered for the reading “kojin”, “deceased” is displayed in FIG.
As shown in (d), the area is registered in the notation 1 and the area of the notation 2 is not taken.

Next, when "individual" is selected for "kojin", the number of homonyms for the reading of "kojin" is two, which is less than seven (step 107).
As a result (step 108), "individual" is registered in the notation 1, and "deceased" is deleted from the priority word dictionary 6b.

As described above, according to the present embodiment, in the case of a reading having a small number of homophones, it is registered only in the notation 1, and in the case of a word having a large number of homophones, it is registered in the notation 1 and the notation 2. Many words, that is, words having a large number of candidates and a high possibility of performing re-conversion can be registered in the priority word dictionary 6b. Also,
In the case of a reading with a small number of homophone words, only the notation 1 is registered, so that the storage area of the priority word dictionary 6b can be used effectively. In the present embodiment, the number of homonyms that can be registered in the priority word dictionary 6b is divided into two types, notation 1 and notation 2. However, the number of homonyms that are read in the basic dictionary 3 can be increased to a greater number. May be divided.

The notation 1 of the preferred word dictionary 6b of the present embodiment.
9 shows an example in which the kanji notation is stored as shown in FIG. 9, but instead of storing the kanji notation, for example, the address of the kanji notation in the basic dictionary 3 may be used. A kanji notation that can be specified may be stored.

Embodiment 4 FIG. FIG. 11 is an overall configuration diagram of a fourth embodiment of a document input device according to the present invention. Components similar to those of the first embodiment shown in FIG. In FIG. 11, reference numeral 6a denotes a preferred word dictionary similar to that of the second embodiment shown in FIG. 4, 8a denotes a preferred word dictionary search means similar to the second embodiment shown in FIG. The operation amount totaling means 13 for measuring the total of the operation amounts input up to the selection, determines the importance of the kanji-mixed notation selected by the user from the total of the operation amounts, and selects the candidate selected by the user according to the determined importance. Is an operation amount priority determining means for changing the recording period in which is stored in the priority word dictionary 6a.
2c is composed of a basic dictionary search unit 4, a phrase generation unit 5, a priority word dictionary search unit 8a, an operation amount totalization unit 12, and an operation amount priority determination unit 13, and is described later according to a control program written in a ROM or the like. This is a control unit that performs a process to be performed.

FIG. 12 shows a preferred word dictionary 6 in this embodiment.
FIG. 6 is an explanatory diagram of a, similar to the preferred word dictionary 6a of Embodiment 2 shown in FIG.
The count indicating the age of registration is stored as shown in FIG. In this embodiment, when the number of reconversions is 5 or more, the operation amount is determined to be significant and important, and is regarded as an important word of rank A. In addition, the number of retransformations less than 5 is defined as a rank B general term. "Count" indicates the age of the words registered in the priority word dictionary 6a as in the second embodiment, and the larger the value, the older the registration in the priority word dictionary 6a. For example, "this"
"Gong" is the second candidate, and "Item" is the seventh
For the candidate, reading “Kojin”, “deceased” is the second candidate, and for the reading “Kisha”, “train” is the second candidate. Further, it is assumed that there is no word registered as the initial state in the priority word dictionary 6a, and 300 words can be registered.

FIG. 13 is a flowchart showing the operation of this embodiment. The operation of this embodiment will be described with reference to the flowchart of FIG. First, a case in which “Gong” is selected for the reading “Kou” will be described. As in the first embodiment, input of kana (step 131), retrieval of the basic dictionary 3 (step 132), generation of a phrase (step 13)
3), a process of searching the priority word dictionary 6a (step 134). The number of reconversion operations is detected by the operation amount totaling means 12. A variable c is prepared to measure how many times the reconversion key has been pressed (step 135), and the value of c is incremented by one each time the reconversion key is pressed in step 137 (step 138). Since this is the first registration, there is a storage area for registering a new word in the priority word dictionary 6a (step 139), and the previously registered word is deleted (step 14).
0) and increment of the count (step 141)
Is not performed. “Gong” is the second candidate, and the value of c (the number of reconversions) is 1, so N is “NO” in step 142. Is determined, the rank is registered in the priority word dictionary 6a, and the count is registered with the initial value of 0 (step 14).
4). As a result, the state of the priority word dictionary 6a is as shown in FIG.
become that way.

Next, in response to the reading "kojin", "deceased"
Is selected, the same processing as described above is performed up to step 139. Next, in step 141, the priority word dictionary 6a
Is a rank B general word, so the count is increased by two. "Dead people" to register
Is a second candidate, and the value of c (the number of re-conversions) is 1, so N (NO) is determined in step 142, and the “deceased” is determined by the operation amount priority determining unit 13 to be a rank B common word. Then, the rank is registered in the priority word dictionary 6a with the rank B and the initial value is 0 (step 144). As a result,
The state of the priority word dictionary 6a is as shown in FIG.

Next, when “item” is selected in response to the reading “ko”, in step 141, “ko” and “deceased” registered in the priority word dictionary 6a are common words of rank B. Therefore, the count is increased by two each time, and the “term” to be registered is the seventh candidate, and the value of c (the number of re-conversions) is 6, so that Y (YES) is obtained in step 142, and the “term” is Key word of rank A, count is initial value 0
(Step 143). At the time of registration, “Ko” has already been registered in the priority word dictionary 6a by reading “Ko”, but “Ko” is deleted by the processing of step 143 because the new item “Item” is prioritized. As a result, the state of the priority word dictionary 6a is as shown in FIG. In this case, “gong” is deleted, but may be left in the priority word dictionary 6a as a second candidate.

Next, when “train” is selected for the reading “kisha”, in step 141, two counts are given to “deceased” of rank B, and “count” is ranked A. Is incremented by one,
Since “train” is the second candidate, step 144
, The rank is B, the count is 0 as the initial value, and the preferred word dictionary 6
a. As a result, the state of the priority word dictionary 6a is as shown in FIG.

As described above, in this embodiment, as in the second embodiment, when a new word is registered, the count of general words already registered in the priority word dictionary 6a is increased by two, and the count of important words is increased by one. As the number of words increases, the number of words registered in the priority word dictionary 6a eventually exceeds 300, and there is no storage area to be registered.
When the storage area is exhausted, the word having the largest count is deleted in step 140, so that the important word remains in the priority word dictionary 6a for a longer time than the general word. Deletion of the word having the largest count value is performed in the priority word dictionary 6a.
This is a process necessary to effectively use the storage area of the general term. According to the present embodiment, general words with low importance can be deleted before important words.

In addition to the method using the count, even if the number of words to be registered is divided according to the degree of importance, the important words can be left in the priority word dictionary 6a for a long time. As an example, it is assumed that the number of words that appear less than five times is about ten times greater than the number of words that appear when the conversion key is pressed five or more times. FIG.
As shown in the explanatory diagram of the preferred word dictionary 6a, the number of words that can be registered is 300 words, of which 100 words are important words and 20 words
If 0 words are assigned to the registration of general words, the priority word dictionary 6
Since the ratio of important words to general words that can be registered in a is larger, it becomes possible to register the important words longer in the priority word dictionary 6a.

In this embodiment, when the number of re-conversions is 5 or more, the operation amount is determined to be significant and important. However, the number of re-conversions is not limited to five, and the number of re-conversions is not limited to five. You may. Further, in this embodiment, 300 words can be registered in the priority word dictionary 6a. However, the maximum number may be more than 300 words and may be determined by the size of the storage area of the priority word dictionary 6a.

Although the preferred word dictionary 6a according to the present embodiment has an example in which the kanji notation is stored as shown in FIG. 12, instead of storing the kanji notation, for example, the kanji notation in the basic dictionary 3 is used. Or an address that can specify the kanji notation stored in the basic dictionary 3.

Embodiment 5 FIG. FIG. 14 is an overall configuration diagram of a fifth embodiment of a document input device according to the present invention. Components similar to those of the first embodiment shown in FIG. In FIG. 14, reference numeral 6d denotes a priority word dictionary for recording candidates selected by the user, reference numeral 8d denotes a priority word dictionary search unit for searching the priority word dictionary 6d from the reading of the phrases separated by the phrase generation unit 5, and reference numeral 12 denotes a fourth embodiment. The operation amount totaling means for measuring the total of the operation amounts input by the user until the user selects the target kanji mixture candidate as in the case of the operation amount totalization means 12, and 14 is the operation amount The operation amount registration homonym number determining means for changing the number of homonyms to be registered in the priority word dictionary 6d according to the determined importance. 2d is a basic dictionary search means 4, a phrase generation means 5, a priority word dictionary search means 8d, and an operation amount totaling means 1
2 and an operation amount registration homophone number determination means 14,
According to the control program written in ROM etc.,
This is a control unit that performs processing described below.

FIG. 15 shows a preferred word dictionary 6 in this embodiment.
It is explanatory drawing of d. As shown in FIG.
“d” has two types of notations for reading a phrase, notations 1 and 2, and ranks 1 and 2 indicating the importance of each notation. As in the case of the fourth embodiment, the rank in this embodiment is defined as a keyword A rank when the number of reconversions is 5 or more and a general word rank B when less than 5 times. The notation 1 corresponds to the rank 1 and the notation 2 corresponds to the rank 2, and the notation 1 is displayed as the first candidate and the notation 2 is displayed as the second candidate for the reading. As an example, the pronunciation "ko" is replaced by
It will be selected in the order of “Gong”, “A” and “Gong”. In the basic dictionary 3, “Gong” is the second candidate, “Item” is the seventh candidate,
“A” is the eighth candidate. It is also assumed that there is no word registered as the initial state in the priority word dictionary 6a.

FIG. 16 is a flowchart showing the operation of this embodiment. The operation of this embodiment will be described with reference to the flowchart of FIG. First, a case where “item” is selected for the reading “ko” will be described. As in the fourth embodiment, input of kana (step 161), retrieval of the basic dictionary 3 (step 162), generation of a phrase (step 16)
3) Search the priority word dictionary 6d (step 164), display candidates (step 166), and count the number of reconversions (step 168). Since "item" is the first entry in the priority word dictionary 6d, and no word corresponding to the reading "ko" is registered in the priority word dictionary 6d, it is determined to be N (NO) in step 169. Next, the manipulated variable homophone number determination means 14 determines that the term is an important word of rank A because the term has been reconverted at least seven times in the seventh candidate, and the priority word dictionary 6d
Is registered in the notation 1 and A is registered in the rank 1 (step 176). As a result, the state of the priority word dictionary 6d is as shown in FIG.
(A).

Next, when "Kou" is selected for the reading "Kou", the same processing as described above is performed up to step 168. This time, since only one word “item” is registered in the priority word dictionary 6d with the reading “ko”, Y (YES) is determined in step 169 and step 1 is performed.
It becomes Y (YES) at 70. Next, since “term” in the priority word dictionary 6d is an important word of rank A, Y (YES) is determined in step 174, and
The item is registered in notation 1 as a candidate, and “item” is shifted to notation 2 and rank 2 as a second candidate. It should be noted that the number of reconversions for obtaining “gong” is one, which is less than five, so that the rank of “gong” is stored in rank 1 as B. As a result, the priority word dictionary 6d is as shown in FIG.

Next, when "Ko" is selected for the reading "Kou", the same processing as described above is performed up to step 168, and the word corresponding to the reading "Kou" is stored in the priority word dictionary 6d. Are registered as “Gong” and “Item”, the result is Y (YES) in Step 169, and Step 1
It becomes N (NO) at 70. Since “Gong” is different in rank from B and “Item” is different in rank from A, Y (Y
ES), and the “gong” of the notation 1 with low importance is deleted (step 172), and the new word “a” is registered in the notation 1 (step 176). Since the number of reconversions to obtain “A” is 7 times and 5 or more, the rank of “A” is A, and as a result, the priority word dictionary 6d is shown in FIG.
become that way.

Next, for the reading "KOU", "KON" is selected again. The same processing proceeds up to step 168, and two words corresponding to the reading “ko” are registered in the priority word dictionary 6d, “Ko” and “item”, so that Y (YES) is obtained in step 169, N at step 170
(NO). Since “A” and “item” have the same importance, N (NO) is obtained in step 171 and the priority word dictionary 6
The "term" of the notation 2, which is an old word registered in d, is deleted (step 173), "gong" is newly registered in the notation 1 as a first candidate, and "A" ranks with the notation 2 as a second candidate. 2 (step 176). Since the number of reconversions to obtain "Gong" is one and less than five,
The rank of “K” is B, and as a result, the priority word dictionary 6d
Is as shown in FIG.

Note that only one general word of rank B is registered in the priority word dictionary 6d, and the word registered this time is also rank B
In step 169, Y (YES), in step 170, Y (YES), and in step 174, N (N
O), and at step 175, the notation 1 of the priority word dictionary 6d
Is deleted, and a new word is registered in step 176.

As described above, in this embodiment, the preferred word dictionary 6d
If the number of homonyms already registered is 0, a new word is registered in notation 1. If the number of homophones is 1, if the registered word is a general word, delete the registered word and then register a new word in the notation 1,
If the registered word is an important word, the registered word is shifted to notation 2 and a new word is registered in notation 1. If the number of homophones is 2, if both registered words are important words, delete the old registered word, shift the remaining words to notation 2, register the new word in notation 1, and register the new word in notation 1. If the importance is different, a general word having a low importance is deleted, and the remaining word is shifted to the notation 2, and a new word is registered in the notation 1. Therefore, according to the present embodiment,
Since important words whose reconversion times are 5 or more are registered in both the notation 1 and the notation 2, many important words can be registered in the priority word dictionary 6d. In addition, a general word whose reconversion is less than 5 is registered only in the notation 1, and an old registered word or a word of low importance is deleted first, so that the storage area of the priority word dictionary 6d is effectively used. Available.

In the present embodiment, the importance is divided into two, ie, the number of times of reconversion is 5 or more and less than 5, but the importance may be further finely divided depending on the amount of operation. Further, in the present embodiment, the number of homonyms that can be registered in the priority word dictionary 6d is divided into two types, notation 1 and notation 2. However, the number of homonyms that are read in the basic dictionary 3 can be further increased. May be divided.

In addition, the notation 1 of the preferred word dictionary 6d of this embodiment
15 shows an example in which the kanji notation is stored as shown in FIG. 15, but instead of storing the kanji notation, for example, the address of the kanji notation in the basic dictionary 3 may be used.
May be stored.

Embodiment 6 FIG. FIG. 17 is an overall configuration diagram of a sixth embodiment of the document input device according to the present invention. Components similar to those of the first embodiment shown in FIG. In FIG. 17, 6e is a preferential word dictionary for recording candidates selected by the user, 8e is a preferential word dictionary search means for searching the preferential word dictionary 6e from reading of phrases separated by the clause generating means 5, and 15 is selected. It is a connection relation determining means for determining whether or not there is a connection relation between candidate words from the kanji notation attribute information registered in the basic dictionary 3 and registering the selected word including the connection relation in the priority word dictionary 6e. . 2e is composed of a basic dictionary search means 4, a phrase generation means 5, a priority word dictionary search means 8e, and a connection relation judgment means 15,
It is a control unit that performs processing described below in accordance with a control program written in, for example,.

FIG. 18 shows the preferred word dictionary 6 in this embodiment.
It is explanatory drawing of e. In the preferred word dictionary 6e, reading of a phrase,
The notation, and the preceding and succeeding words for this notation are stored as shown in FIG. When it is determined that two words have a connection relationship, the connected word is registered in one notation. Also,
The pre-word is the word before the connected word, and the post-word is the word after the connected word. As an example, a case will be described in which the notation "mysterious phenomenon" is obtained for the reading "kaigensho". Note that the basic dictionary 3 does not include the word “mysterious phenomenon”, and the prefix “mystery” and “kai”
It is assumed that nouns such as “solution”, “floor”, “shell”, and “lower” are registered as words having higher priority than “monster”, and that the word “phenomenon” (noun) is also registered. Also,
A word consisting of a prefix and a noun is connected. Also,
It is assumed that there is no word registered as the initial state in the priority word dictionary 6e.

FIG. 19 is a flowchart showing the operation of this embodiment. The operation of this embodiment will be described with reference to the flowchart of FIG. When the user enters “Kaigensho” (step 191), as in the first embodiment,
Retrieval of basic dictionary 3 (step 192), generation of phrase (step 193), retrieval of preferential word dictionary 6e (step 19)
4), candidates are displayed (step 195), and re-conversion (step 196) is performed, and the phrase is divided into "kai" and "gensho". Here, "kai" for "kai"
However, suppose that "phenomenon" is displayed for "gensho". Next, in step 196, the re-conversion key is pressed several times, "kai" is displayed for "kai", and "gensho" is fixed as "phenomenon". Next, the connection relation judging means 15 knows that the part of speech of "monster" is a prefix and the part of speech of "phenomenon" is a noun from the basic dictionary 3, and judges that there is a connection relation between the two words (step 197). ),
Register as a connection word (step 198).

As a result, FIG.
As shown in (a), the reading “Kaigensho” is registered with the notation “mysterious phenomenon”, the previous word “mysterious”, and the subsequent word “phenomenon”. After that, if "Kaigensho" is entered,
“Mysterious phenomenon” is displayed as a first candidate. At this time, since there is no word registered under the heading of “kai”, “kai” is not given priority to “kai” other than “kaigensho” and is displayed in the priority order of the basic dictionary 3.

If it is determined in step 197 that there is no connection, the words are registered in the priority word dictionary 6e in word units. For example, when “lower” (noun) and “phenomenon” (noun) are selected for “kaigensho”, step 197
It is determined that there is no connection between the noun and the noun,
As shown in FIG. 18 (b), the heading “KAI” indicates “lower”.
Is registered (step 199), and with respect to the candidate corresponding to the reading “kai”, “lower” is given priority.

In the present embodiment, the connection relation is a prefix and a noun, but a noun and a suffix, or a noun and a noun that do not sandwich an adjunct, may be subjected to the connection word learning. Although the preferred word dictionary 6e according to the present embodiment has an example in which the kanji notation is stored as shown in FIG. 18, instead of storing the kanji notation, for example, the address of the kanji notation in the basic dictionary 3 may be used. It is also possible to store a kanji notation stored in the basic dictionary 3.

[0082]

As described above, according to the present invention, the recording period to be recorded in the priority word dictionary according to the importance determined from the part of speech information of the kanji notation registered in the selected basic kanji mixed notation dictionary. As it changes, words with high importance can be left in the preferred word dictionary for a long time. Furthermore, since words with low importance are deleted first, the storage area of the priority word dictionary can be used effectively.

[0083]

According to another aspect of the invention, the recording period for recording in the priority word dictionary is changed according to the importance determined from the attribute information of the kanji notation registered in the basic dictionary. Can be left in the preferred word dictionary.
Furthermore, since words with low importance are deleted first, the storage area of the priority word dictionary can be used effectively.

According to another invention, the number of homophones recorded in the priority word dictionary is changed according to the importance determined from the number of homophones for the kana registered in the basic dictionary.
Many words with high importance can be left in the preferred word dictionary. Furthermore, since words with low importance are deleted first, the storage area of the priority word dictionary can be used effectively.

According to another aspect of the present invention, the importance determined from the total of the operation amounts input until the target kanji mixture candidate is selected and the selected candidate are recorded in the priority word dictionary,
Since the recording period for recording in the priority word dictionary is changed according to the recorded importance, words having high importance can be left in the priority word dictionary for a long time. Furthermore, since words with low importance are deleted first, the storage area of the priority word dictionary can be used effectively.

According to another aspect of the present invention, the importance determined from the total of the operation amounts input until the target kanji mixture candidate is selected and the selected candidate are recorded in the priority word dictionary,
Since the number of homologous words recorded in the priority word dictionary is changed according to the recorded importance, many words having high importance can be left in the priority word dictionary. Furthermore, since words with low importance are deleted first, the storage area of the priority word dictionary can be used effectively.

[0088]

[Brief description of the drawings]

FIG. 1 is an overall configuration diagram of a document input device according to a first embodiment of the present invention.

FIG. 2 is an explanatory diagram of a preferred word dictionary used in Embodiment 1 of the present invention.

FIG. 3 is a flowchart illustrating an operation of the document input device according to the first embodiment of the present invention.

FIG. 4 is an overall configuration diagram of a document input device according to a second embodiment of the present invention.

FIG. 5 is an explanatory diagram of a preferred word dictionary used in Embodiment 2 of the present invention.

FIG. 6 is a flowchart showing an operation of the document input device according to the second embodiment of the present invention.

FIG. 7 is an explanatory diagram of a preferred word dictionary used in Embodiment 2 of the present invention.

FIG. 8 is an overall configuration diagram of a document input device according to a third embodiment of the present invention.

FIG. 9 is an explanatory diagram of a preferred word dictionary used in Embodiment 3 of the present invention.

FIG. 10 is a flowchart showing an operation of the document input device according to the third embodiment of the present invention.

FIG. 11 is an overall configuration diagram of a document input device according to a fourth embodiment of the present invention.

FIG. 12 is an explanatory diagram of a preferred word dictionary used in Embodiment 4 of the present invention.

FIG. 13 is a flowchart illustrating an operation of the document input device according to the fourth embodiment of the present invention.

FIG. 14 is an overall configuration diagram of a document input device according to a fifth embodiment of the present invention.

FIG. 15 is an explanatory diagram of a preferred word dictionary used in Embodiment 5 of the present invention.

FIG. 16 is a flowchart showing an operation of the document input device according to the fifth embodiment of the present invention.

FIG. 17 is an overall configuration diagram of a document input device according to a sixth embodiment of the present invention.

FIG. 18 is an explanatory diagram of a preferred word dictionary used in Embodiment 6 of the present invention.

FIG. 19 is a flowchart illustrating an operation of the document input device according to the sixth embodiment of the present invention.

FIG. 20 is an explanatory diagram for explaining an operation of a conventional document input device.

[Description of sign]

 DESCRIPTION OF SYMBOLS 1 Input means 3 Basic dictionary 4 Basic dictionary search means 5 Clause generation means 6 Priority word dictionary 7 Phrase information extraction means 8 Priority word dictionary search means 9 Display means 10 Attribute priority determination means 11 Basic registration homonym number determination means 12 Total operation amount Means 13 Operation amount priority determination means 14 Operation amount registration homophone number determination means 15 Connection relation determination means

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Masato Nishihara 5-1-1, Ofuna, Kamakura-shi Mitsubishi Electric Corporation Personal Information Equipment Development Laboratory (56) References JP-A-3-167659 (JP, A JP-A-60-57425 (JP, A) JP-A-3-48968 (JP, A) JP-A-60-57425 (JP, U) (58) Fields investigated (Int. Cl. ⁷ , DB name) G06F 17/22

Claims (5)

(57) [Claims] 1. An input means for inputting a Japanese sentence with a kana,
A basic dictionary storing kanji notation corresponding to kana and attribute information of the notation; a basic dictionary search means for searching the basic dictionary with a partial character string of a kana character string input by the input means; and a basic dictionary search A phrase generating means for separating an input kana character string into Japanese phrases from information retrieved by the means to generate a kanji-mixed notation, a display means for displaying kanji-mixed notation candidates, and inputting from the candidates displayed on the display means. A preferred word dictionary in which the independent word of the candidate selected by the means is recorded; and a preferred word dictionary search means for searching the preferred word dictionary from reading of the independent word in the phrase separated by the clause generating means. the document input device, the candidate selected by said input means as well as recorded in the priority dictionary, to determine the importance of the part of speech of the candidate, its weight Priority term determining means for changing a reference of a period recorded in the priority word dictionary in accordance with the degree of search, wherein the priority word dictionary search means searches the priority word dictionary from reading of phrases separated by the phrase generating means A document input device characterized in that the latest use candidate based on the importance is given priority. 2. An input means for inputting a Japanese sentence with a kana,
Kanji notation corresponding to kana and attribute information of this notation are stored.
Basic dictionary and kana sentences input by the input means
Basic dictionary search that searches the above basic dictionary with a substring of a character string
Search means and information searched by the basic dictionary search means
Input kana character string is divided into Japanese phrases and mixed with kanji
Clause generation means for generating notation, and candidates for notation mixed with kanji
And a candidate displayed on the display means.
Records the independent word of the candidate selected by the above input means
And the preferred word dictionary to be separated
Search the preferred word dictionary from the reading of independent words in
Document input device provided with preferred word dictionary search means
The candidate selected by the input means
Record in the dictionary and register it in the above basic dictionary.
Importance is determined from the number of homonyms
Of the period recorded in the above-mentioned preferred word dictionary according to the importance of
A priority determining means for changing the priority is provided, and the priority word dictionary search is performed.
Means for reading the phrase separated by the phrase generating means.
From the priority word dictionary, and
A sentence characterized by giving priority to new use candidates
Input device. 3. The priority determination means determines the importance from the number of homophones for the kana registered in the basic dictionary, and changes the number of homophones recorded in the priority word dictionary according to the determined importance. 3. The document input device according to claim 2, wherein said document input device comprises basic registered homophone number determining means for changing a reference of a period recorded in said priority word dictionary. 4. An input means for inputting a Japanese sentence with a kana,
Kanji notation corresponding to kana and attribute information of this notation are stored.
Basic dictionary and kana sentences input by the input means
Basic dictionary search that searches the above basic dictionary with a substring of a character string
Search means and information searched by the basic dictionary search means
Input kana character string is divided into Japanese phrases and mixed with kanji
Clause generation means for generating notation, and candidates for notation mixed with kanji
And a candidate displayed on the display means.
Records the independent word of the candidate selected by the above input means
And the preferred word dictionary to be separated
Search the preferred word dictionary from the reading of independent words in
Document input device provided with preferred word dictionary search means
The candidate selected by the input means
In addition to recording in a dictionary, before the above candidate is selected
Measures the total amount of input operation, and checks whether the total
The above-mentioned importance is determined from the
The selected candidates are recorded in the above preferred word dictionary, and the recorded importance is recorded.
Change the recording period to record in the above preferred word dictionary according to
The priority word dictionary search means is provided.
Is the reading of the phrase separated by the above phrase generation means?
Search the above preferred word dictionary, and use the latest
Document input characterized by giving priority to
Power device. 5. The priority determination means, wherein the operation amount totalization means for measuring the total of the operation amounts inputted until the target kanji mixture candidate is selected, and the priority is determined from the total of the operation amounts. The determined importance and the selected candidate are recorded in the priority word dictionary, and the number of homophone words to be recorded in the priority word dictionary is changed according to the recorded importance. 5. The document input device according to claim 4, wherein said document input device comprises an operation amount registration homophone number determining means for changing a reference of said operation amount.