JP5326945B2 - Character input support device, program, and character input support method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a character input support device ranking a character string having a frequently used meaning in the higher rank of input prediction candidates by summarizing frequencies of use for each group of character strings having the same meaning. <P>SOLUTION: The character input support device includes: an independent word extraction part 205 extracting an independent word from partial character string data; a first detection part 206 selecting partial character string data of a processing target from the partial character string data and detecting other partial character string data including the independent word included in the partial character string data of the processing target; a first selection part 207 selecting partial character string data from the detected other partial character string data and the partial character string data of the processing target according to the length or an appearance frequency of the partial character string data and storing the selected partial character string data as an input prediction candidate in a RAM 103; and a frequency recalculation part 204 calculating a total value of the appearance frequency of the partial character string data of the processing target and that of the detected other partial character string data and recording the calculated total value in the RAM 103 as an application frequency of the input prediction candidate. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to a character input support device, a program, and a character input support method.

  2. Description of the Related Art Character input support devices that support user keyboard input are known. The character input support device stores an input history of a character string input from the keyboard in the past in the storage unit, predicts a character string to be input next with reference to the input history, and determines a predicted input prediction candidate. It is a device that displays on the display unit. When one of the input prediction candidates displayed on the display unit is specified by the user, the specified character string is input to the input area. This type of technology is disclosed in Patent Documents 1 and 2.

Japanese Patent No. 3933952 JP 2007-334534 A

  The same contents may be expressed by different character strings. For example, “blood pressure: normal”, “blood pressure is normal”, and “blood pressure normal” all indicate the same content. In the conventional character input support device, the frequency is totaled for each character string, and is not tabulated according to the meaning and contents of the character string. For this reason, although the frequency of use is high, it is difficult to rank the content expressed by a plurality of character strings, or the character string representing this content at the top of the input prediction candidates, which causes a decrease in user convenience. It was.

  The present invention has been made in view of the above circumstances, and by summing up the frequency of use for each group of character strings having the same meaning, character strings having a high usage frequency are ranked higher in the input prediction candidates. It is an object to provide a character input support device, a program, and a character input support method that can be ranked.

The character input support device disclosed in this specification receives input by an input receiving unit, analyzes a plurality of character string data stored in a storage unit, and converts each character string data into a word included in each character string data. Alternatively, conversion means for converting to a plurality of partial character string data including at least two clauses, and a frequency count for counting the appearance frequency of the partial character string data converted by the conversion means in the plurality of character string data Means for extracting independent words from the partial character string data converted by the converting means; and selecting the partial character string data to be processed from the partial character string data converted by the converting means Then, referring to the extraction result of the independent word extraction unit, other partial character string data including the independent word included in the partial character string data to be processed is converted into the partial character string data to be processed. A first character detection unit that detects partial character string data having the same meaning as the first character, a partial character string among the other partial character string data detected by the first detection unit and the partial character string data to be processed The partial character string data is selected according to the length of the data or the appearance frequency counted by the frequency counting means, and the selected partial character string data is stored in the storage means as an input prediction candidate that supports the input of the character string. Calculating a total value of appearance frequencies of the first selection means to be processed, the partial character string data to be processed, and the other partial character string data detected by the first detection means, and the calculated total value is the input Based on the frequency of use stored in the storage means when receiving the input of character string data by the frequency calculation means to be stored in the storage means as the use frequency of the prediction candidate and the input reception means, Elected partial character string data containing the character string data received power, and a selection means to be displayed on the display means selected by the partial character string data.
The character input support device disclosed in this specification extracts an independent word from partial character string data, and detects partial character string data including the same independent word as partial character string data having the same meaning. In addition, by selecting one of the partial character string data having the same meaning and using the sum of the frequency of use of the character string data having the same meaning as the frequency of use of the selected character string data, the portion having the same meaning Usage frequency is aggregated for each group of character string data. Therefore, it is possible to rank a character string having a high frequency of use at a higher rank than the input prediction candidates.

  According to the character input support device disclosed in the present specification, by summing up the frequency of use for each group of character strings having the same meaning, the character string having a high usage frequency is ranked higher in the input prediction candidates. Can be ranked.

It is a block diagram which shows an example of a structure of a computer apparatus. It is a figure which shows an example of a structure of the functional block comprised in a control part by program control. It is a block diagram which shows an example of a structure of the input prediction candidate DB production | generation part shown in FIG. It is a figure which shows an example of the recording information recorded on RAM, and is a figure which shows a mode that the phrase string which divided | segmented the character string was recorded for every number of phrases. (A) is a figure which shows the count result which counted the appearance frequency which appears in the character string data with which 2 phrase strings were recorded as input history, and (B) was recorded with 3 phrase strings as input history. It is a figure which shows the count result which counted the appearance frequency which appears in character string data, (C) is a figure which shows the result of having sorted the 2 phrase string shown to (A) by using the first character as a key, ( (D) is a diagram showing a result of sorting the three character strings shown in (B) using the first character as a key. It is a figure which shows an example of the recording information recorded on RAM, and is a figure which shows an example of the recording information finally recorded. It is a flowchart which shows the whole process sequence of a control part. It is a flowchart which shows the detailed procedure of step S3 shown in FIG. It is a flowchart which shows the detailed procedure of step S4 shown in FIG. It is a flowchart which shows the detailed procedure of step S5 shown in FIG. It is a figure which shows the 1st example of a display of the input prediction candidate displayed on a display. It is a figure which shows the 2nd example of a display of the input prediction candidate displayed on a display. It is a figure which shows the 3rd example of an input prediction candidate displayed on a display, (A) is an example of the screen which requests | requires the input of the narrowing word which further narrows down the character string to display, (B), It is a figure which shows the example which displayed the upper input prediction candidate according to the narrowed-down word input. It is a figure which shows the 4th example of a display of the input prediction candidate displayed on a display, (A) is the example which displayed the character string input by the user, and the narrowing word which narrows down an input prediction candidate, ( B) and (C) are diagrams showing display examples of input prediction candidates when one of the narrowed words is selected.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

First, the configuration of a computer device 1 equipped with the character string input support device of the present invention will be described with reference to FIG.
The computer apparatus 1 includes a control unit 100, a storage unit 104, a graphic board 111, interfaces (hereinafter abbreviated as I / F) 113 and 114, and a network interface (hereinafter abbreviated as network I / F) 118. And a display 112 and an operation unit 115. The control unit 100 includes a CPU 101, a ROM 102, and a RAM 103. The operation unit 115 includes a keyboard 116 and a mouse 117.

  First, the control unit 100 will be described. A CPU (Central Processing Unit) 101 controls the operation of the entire apparatus. A ROM (Read Only Memory) 102 stores in advance a control program for operating the computer device 1 as a character input support device and other various programs. A RAM (Random Access Memory) 103 stores data used by the CPU 101 for calculation, data of calculation results, and the like.

The storage unit 104 is a storage device such as an HDD (Hard Disk Drive) provided in the computer apparatus 1. The storage unit 104 stores a character string history database (hereinafter, the database is referred to as DB) in which a history of character strings input from the keyboard in the past is recorded. In addition, the storage unit 104 determines candidates to be input next based on the Kana-Kanji conversion candidate DB in which a dictionary for converting input Kana characters into Kanji is registered or the input history of character strings input in the past. An input prediction candidate DB for prediction is stored. Furthermore, the storage unit 104 stores a dictionary necessary for morphological analysis, a synonym dictionary for extracting synonyms of independent words, and a dictionary in which symbols representing the meaning of words are registered.
The storage unit 104 is not limited to the HDD, but is a computer-readable flexible disk (FD), DVD (Digital Versatile Disc), DVD-RAM, CD-R (Recordable) / RW (ReWritable) magneto-optical. It may be a portable storage medium such as a disk or an IC card.

  The graphic board 111 controls display of various types of information on the display 112. The I / F 113 sends the character information received from the keyboard 116 to the CPU 101. The I / F 114 sends input information received by the mouse 117 to the CPU 101. The network I / F 118 transmits / receives communication data to / from an external device connected to the network 119.

  The CPU 101, ROM 102, RAM 103, storage unit 104, graphic board 111, I / F 113, 114, and network I / F 118 are connected to the system bus 105. Accordingly, the CPU 101 accesses the ROM 102, RAM 103, and storage unit 104, displays various information on the display 112 via the graphic board 111, inputs information by operating the keyboard 116 and mouse 117, and the network I / F 118. Communication data can be transmitted / received via the network.

Next, functional blocks of the control unit 100 realized by program control will be described with reference to FIGS. 2 and 3. The functional block is a block in which processing realized by cooperation of hardware such as the CPU 101 and the RAM 103 included in the control unit 100 and the control program stored in the ROM 102 is grouped for each predetermined function.
As shown in FIG. 2, the control unit 100 includes an input prediction candidate DB creation unit 200, a kana-kanji conversion processing unit 250, an input prediction candidate search unit 300, and a display control unit 350.

  The input prediction candidate DB creation unit 200 creates an input prediction candidate DB stored in the storage unit 104. In the storage unit 104, a history of character strings input by the user from the keyboard 116 in the past is recorded. The input prediction candidate DB creation unit 200 refers to the character string history and selects a character string to be displayed in the display area of the display 112 as an input prediction candidate. The selected character string is recorded in the input prediction candidate DB as an input prediction candidate. Details of the input prediction candidate DB creation unit 200 will be described later with reference to FIG.

  The kana-kanji conversion processing unit 250 refers to the kana-kanji conversion candidate DB stored in the storage unit 104 and performs kana-kanji conversion processing for the characters that have been accepted by the keyboard 116. When the kana-kanji conversion processing unit 250 receives input of kana characters from the keyboard 116, the kana-kanji conversion candidate DB refers to the kana-kanji conversion candidate DB and selects a conversion candidate for converting the input kana character into kanji. The selected conversion candidate is displayed on the display 112 by display control of the display control unit 350.

  The input prediction candidate search unit 300 refers to the above-described input prediction candidate DB, and selects an input prediction candidate of a character string that has been input with the keyboard 116. The selected input prediction candidate is output from the input prediction candidate search unit 300 to the display control unit 350 and displayed on the display 112 under the control of the display control unit 350.

  The display control unit 350 causes the display 112 to display the conversion candidate selected by the kana-kanji conversion processing unit 250 and the input prediction candidate selected by the input prediction candidate search unit 300. A display example of the conversion candidate and the input prediction candidate on the display 112 will be described later.

  Next, the details of the input prediction candidate DB creation unit 200 will be described with reference to FIG. The input prediction candidate DB creation unit 200 includes a conversion unit 201, a frequency counting unit 202, a third detection unit 203, a frequency recalculation unit 204, an independent word extraction unit 205, a first detection unit 206, a first selection unit 207, and synonyms. An extraction unit 208, a second detection unit 209, and a second selection unit 210 are provided.

The conversion unit 201 inputs and analyzes the past keyboard input history registered in the character string history DB 401. The character string history DB 401 is stored in the storage unit 104. The conversion unit 201 refers to the morphological analysis dictionary stored in the dictionary DB 402 and performs morphological analysis of each character string recorded as the input history. The dictionary DB 402 is also stored in the storage unit 104. The conversion unit 201 divides each character string into phrases based on the result of morphological analysis. Symbols such as “:” and “?” Are also divided as one phrase. Next, the conversion unit 201 converts a plurality of clauses (hereinafter referred to as a phrase string. The phrase string corresponds to a partial character string of the present invention) from a character string divided into phrases into two phrases, three phrases, and the like. Take out. The number of clauses to be extracted is limited to the consecutive (connected) clauses in the character string, and the maximum number of clauses to be extracted is the same as the number of clauses included in the target character string. For example, it is assumed that the character string is “blood pressure is slightly higher than at the time of last week's visit”. In this case, the character string includes 6 clauses (last week), (at the time of consultation), (compared), (blood pressure), (a little), (higher). Therefore, as a phrase string including two phrases (hereinafter, referred to as a two-phrase string), the conversion unit 201 performs (in the last week / at the time of consultation), (compared at the time of consultation), (compared to / blood pressure), ( The blood pressure can be taken out (a little) and (a little / high). Similarly, as a phrase string including three phrases (hereinafter, referred to as a three-phrase string), the conversion unit 201 (the blood pressure), (when compared to / when compared with the last week), Compared to this, four blood pressures can be taken out: (blood pressure is a little) and (blood pressure is a little / high) Thereafter, the same procedure is repeated, and the conversion unit 201 extracts 2 to 6 phrase strings. In this specification, “/” is used as a symbol indicating a break between clauses. FIG. 4 shows a result of dividing four character strings of “blood pressure: normal”, “blood pressure is normal”, “blood pressure normal”, and “blood pressure: abnormal” into phrases, and a phrase string is created from the divided phrases. Results are shown. The conversion unit 201 sends the generated phrase string data to the third detection unit 203, the frequency counting unit 202, the independent word extraction unit 205, and the synonym extraction unit 208. Further, the conversion unit 201 sends the result of the morphological analysis of the character string to the independent word extraction unit 205. Further, the conversion unit 201 stores the generated phrase string in the RAM 103.
In this description, the conversion unit 201 is described as converting a character string into a phrase string, but may be divided into word strings in units of words. The word string is a group of a plurality of words that are continuous (connected) in the character string, like the phrase string.

  The frequency counting unit 202 acquires the phrase string extracted from the character string by the conversion unit 201 and acquires the character string of the input history from the character string history DB 401. The frequency counting unit 202 counts the appearance frequency of each phrase string acquired from the conversion unit 201 in the character string recorded as the input history in the character string history DB 401. The frequency counting unit 202 records the counted appearance frequency of each phrase string in the RAM 103. FIG. 5A shows a count result obtained by counting the appearance frequency of the two-phrase string shown in FIG. FIG. 5B shows a count result obtained by counting the appearance frequency of the three-phrase string shown in FIG. FIG. 5C shows a sorting result obtained by sorting the two phrase strings by the first character. FIG. 5D shows a sorting result obtained by sorting the three phrase strings by the first character. The frequency counting unit 202 outputs the appearance frequency of each phrase string to the third detection unit 203, the first detection unit 206, and the second detection unit 209.

The third detection unit 203 acquires the phrase string extracted from the character string from the conversion unit 201, and acquires the appearance frequency data of each phrase string from the frequency counting unit 202. First, the third detection unit 203 selects a phrase string to be processed from the phrase strings acquired from the conversion unit 201. The third detection unit 203 includes a clause string that includes more clauses than the number of clauses included in the selected clause string to be processed, and includes another clause string that includes a clause included in the clause string to be processed. To detect. For example, in the example shown in FIG. 4, “blood pressure /:” in the 2-phrase string is included in “blood pressure /: / normal” in the 3-phrase string and “blood pressure /: / abnormal” in the 3-phrase string. Therefore, when “blood pressure /:” is selected as the phrase string to be processed, the third detection unit 203 sets “blood pressure /: / normal” and “blood pressure /: / abnormal” in the three phrase strings as the other phrase strings. Is detected as another phrase string.
When the third detection unit 203 detects another phrase string, the third detection unit 203 updates the recorded information in the RAM 103. FIG. 6 shows an example of record information recorded in the RAM 103. In the RAM 103, an identification number for identifying a phrase string, a phrase string, the frequency of occurrence of the phrase string, the meaning represented by the phrase string, summary information, and a deletion flag are recorded in association with each other. The meaning of the phrase string will be described later. The summary information is information indicating how the corresponding phrase string is processed.
When detecting the other phrase string, the third detection unit 203 records a deletion flag indicating that the phrase string to be processed is excluded from the input prediction candidates in the record column of the phrase string to be processed. In addition, the third detection unit 203 records, as summary information of the phrase strings to be processed, that the phrase strings to be processed are grouped into other detected phrase strings. In addition, the third detection unit 203 records that the processing target phrase strings are collected as summary information of other detected phrase strings. In the example illustrated in FIG. 6, the phrase strings of identification numbers 3 and 6 are combined into the phrase string “blood pressure is normal” of identification number 4. As shown in FIG. 6, a deletion flag may be recorded also in the phrase string of identification number 4, and newly recorded as a phrase string of identification number 7, which is a copy of the phrase string of identification number 4 as it is. In this case, the appearance frequency of the phrase string with the identification number 7 is the total value of the appearance frequencies of the phrase strings with the identification numbers 3, 4, and 6. Details of these rules will be described later with reference to a flowchart.
The third detection unit 203 performs the above processing for all the phrase strings acquired from the conversion unit 201.

  The frequency recalculation unit 204 acquires, from the third detection unit 203, the appearance frequency of other phrase sequences detected by the third detection unit 203 and the appearance frequency of the phrase sequence to be processed. The frequency recalculation unit 204 adds the appearance frequency of another phrase string and the appearance frequency of the phrase string to be processed. The frequency recalculation unit 204 records the addition result in the RAM 103 as the appearance frequency of other character string data. In the example illustrated in FIG. 6, the frequency recalculation unit 204 calculates the total value of the appearance frequencies of the phrase strings with the identification numbers 3, 4, and 6, and records the calculated total value as the appearance frequency of the phrase string with the identification number 7. To do.

The independent word extraction unit 205 acquires from the conversion unit 201 an analysis result obtained by performing a morphological analysis on the character string and a phrase string extracted from the character string by the conversion unit 201.
The independent word extraction unit 205 refers to the acquired morphological analysis result and extracts an independent word from the phrase string. Further, the independent word extraction unit 205 obtains a symbol representing the meaning of the extracted independent word with reference to the dictionary DB 402. In the dictionary DB 402, a dictionary in which independent words and symbols indicating the meanings of the independent words are recorded is registered. The independent word extraction unit 205 refers to this dictionary to obtain a symbol representing the meaning of the independent word. The symbol representing meaning may be another language such as English, or a concept identifier used for an EDR electronic dictionary of the Electronic Dictionary Research (EDR). The concept identifier is a numerical expression of the concept represented by the word. Below, the case where English is used as a symbol showing the meaning of an independent word will be described as an example. For example, when the phrase string “open window” is targeted, the independent words are “window” and “open”, and the symbols representing the meanings of these independent words are “WINDOW” and “OPEN”. . The independent word extraction unit 205 sorts the symbols representing the meanings of the obtained independent words in alphabetical order, and sends the sorting result to the first detection unit 206 and the second detection unit 209. Further, the independent word extraction unit 205 records the sorting result obtained by sorting the symbols representing the meanings of the independent words in the RAM 103. In the example illustrated in FIG. 6, for example, the symbol indicating the meaning of the phrase string “: / abnormal” is “ABNORMAL”.

The first detection unit 206 acquires from the third detection unit 203 a phrase string for which no deletion flag is recorded in the RAM 103. Further, the first detection unit 206 acquires data on the appearance frequency of each phrase string from the frequency counting unit 202. In addition, the first detection unit 206 acquires a phrase string and the independent words included in the phrase string from the independent word extraction unit 205.
The first detection unit 206 selects a phrase string to be processed from the phrase strings acquired from the third detection unit 203 (hereinafter referred to as a phrase string to be processed). Next, the first detection unit 206 recognizes an independent word included in the selected phrase string to be processed based on the information acquired from the independent word extraction unit 205. Further, the first detection unit 206 selects the same independent word as the independent word included in the phrase string to be processed from the phrase strings acquired from the third detection unit 203 based on the information acquired from the independent word extraction unit 205. Detects other included phrase sequences. When other phrase strings are detected, the first detection unit 206 sends the phrase strings to be processed, the detected other phrase strings, and the appearance frequencies of these phrase strings to the first selection unit 207.
The first detection unit 206 performs the above processing for all the phrase strings acquired from the third detection unit 203.

The first selection unit 207 acquires the phrase string to be processed, the other detected phrase strings, and the appearance frequencies of these phrase strings from the first detection unit 206. The first selection unit 207 selects either the phrase string to be processed or the detected other phrase string in accordance with the length of the phrase string or the appearance frequency of the phrase string, and inputs the selected phrase string It records in RAM103 as a prediction candidate.
The first selection unit 207 first compares the length of the acquired phrase string to be processed with the length of another detected phrase string. The first selection unit 207 selects a phrase string having a longer length as an input prediction candidate, and selects a phrase string having a shorter length as a phrase string to be excluded from the input prediction candidates. If the length of the phrase string to be processed is the same as the length of the other detected phrase string, the first selection unit 207 excludes the phrase string having the smaller appearance frequency from the input prediction candidates. Select on a column. In the following, the phrase string excluded from the input prediction candidates is referred to as the excluded phrase string, and the phrase string not excluded from the input prediction candidates is referred to as the remaining phrase string.
The first selection unit 207 rewrites the recording information in the RAM 103 and records a deletion flag indicating that the phrase string is excluded from the input prediction candidates in the recording column of the phrase string to be excluded. In addition, the first selection unit 207 records the fact that the phrase strings to be excluded are collected as summary information of the phrase strings to be excluded. In addition, the first selection unit 207 records that the phrase strings to be excluded are collected as summary information of the phrase strings to be left. The first selection unit 207 sends the frequency of appearance of the phrase string to be processed and the frequency of appearance of other detected phrase strings to the frequency recalculation unit 204.
The frequency recalculation unit 204 adds the appearance frequency of the phrase string to be processed and the appearance frequency of the detected phrase string. The frequency recalculation unit 204 registers the added total frequency in the RAM 103 as the frequency of the phrase string to be left.

  The synonym extraction unit 208 acquires the phrase string and the independent words included in the phrase string from the independent word extraction unit 205. The synonym extraction unit 208 refers to the dictionary DB 402 and obtains an independent word headword acquired from the independent word extraction unit 205. A synonym dictionary is recorded in the dictionary DB 402. In the synonym dictionary, a headword and a word synonymous with the headword are registered. For example, words such as “fatigue”, “fatigue”, and “feeling of fatigue” are registered as synonyms for the heading “fatigue”. Similarly, synonyms “translation” and “translation” are registered for the heading “translation”. Referring to the synonym dictionary, for example, a two-phrase string “I am tired from translation” and a two-phrase string “I am tired from translation” are the same phrase string in the headword. Therefore, it can be determined that “I am tired from translation” and “I am tired from translation” are phrase strings having the same headword, that is, the same meaning. The synonym extraction unit 208 sends the independent word headword to the second detection unit 209.

The second detection unit 209 acquires from the third detection unit 203 a phrase string for which no deletion flag is recorded in the RAM 103. In addition, the second detection unit 209 acquires the appearance frequency of each phrase string from the frequency counting unit 202. In addition, the second detection unit 209 acquires a phrase string, an independent word included in the phrase string, and a symbol representing the meaning of the independent word from the independent word extraction unit 205. In addition, the second detection unit 209 acquires a headword of an independent word from the synonym extraction unit 208.
First, the second detection unit 209 selects a phrase string to be processed from the phrase strings acquired from the third detection unit 203 (hereinafter referred to as a phrase string to be processed). Next, the second detection unit 209 acquires a symbol representing the meaning of the independent word included in the selected phrase string to be processed based on the information acquired from the independent word extraction unit 205. Next, the second detection unit 209 detects, in the phrase sequence acquired from the third detection unit 203, another phrase sequence that includes the same symbol as the symbol representing the meaning of the independent word included in the phrase sequence to be processed. To do. When other phrase strings are detected, the second detection unit 209 sends the phrase string to be processed, the detected other phrase strings, and the appearance frequency of these phrase strings to the second selection unit 210.

Further, the second detection unit 209 acquires a headword of an independent word included in the phrase string to be processed based on the information acquired from the synonym extraction unit 208. Next, the second detection unit 209 detects, from the phrase strings acquired from the third detection unit 203, other phrase strings including the same headword as the independent word headword included in the phrase string to be processed. . When other phrase strings are detected, the second detection unit 209 sends the phrase string to be processed, the detected other phrase strings, and the appearance frequency of these phrase strings to the second selection unit 210.
The second detection unit 209 performs the above process for all the phrase strings acquired from the third detection unit 203.

The second selection unit 210 obtains the phrase string to be processed, other detected phrase strings, and the appearance frequency of these phrase strings from the second detection unit 209. The second selection unit 210 selects one of the phrase string to be processed and another detected phrase string in accordance with the length of the phrase string or the appearance frequency of the phrase string, and inputs the selected phrase string It records in RAM103 as a prediction candidate.
First, the second selection unit 210 compares the acquired length of the phrase string to be processed with the length of another detected phrase string. The second selection unit 210 selects a phrase string having a longer length as an input prediction candidate, and selects a phrase string having a shorter length as a phrase string to be excluded from the input prediction candidates. In addition, when the length of the phrase string to be processed is the same as the length of the other detected phrase string, the second selection unit 210 excludes the phrase string having the smaller appearance frequency from the input prediction candidates. Select on a column. In this case, it is also called a phrase string that excludes a phrase string that is excluded from the input prediction candidates, and is called a phrase string that leaves a phrase string that is not excluded from the input prediction candidates.
The second selection unit 210 rewrites the recording information in the RAM 103 and records a deletion flag indicating that the phrase string is excluded from the input prediction candidates in the recording column of the phrase string to be excluded. In addition, the second selection unit 210 records that the phrase strings to be excluded are collected as summary information of the phrase strings to be excluded. In addition, the second selection unit 210 records that the phrase strings to be excluded are collected as summary information of the phrase strings to be left.
In addition, the second selection unit 210 sends the frequency of appearance of the phrase string to be processed and the frequency of appearance of the other detected phrase strings to the frequency recalculation unit 204. The frequency recalculation unit 204 adds the appearance frequency of the phrase string to be processed and the appearance frequency of the other detected phrase string. The frequency recalculation unit 204 registers the added total frequency as the frequency of the phrase string to be left.

Next, the processing flow of the control unit 100 will be described with reference to the flowcharts shown in FIGS.
First, the control unit 100 receives an input using the keyboard 116, reads the input history of the character string stored in the storage unit 104, and stores it in the RAM 103. Then, the control unit 100 analyzes the character string stored in the RAM 103 one sentence at a time. The control unit 100 performs analysis such as morphological analysis on the character string, and divides the character string into phrases (or words) based on the analysis result (step S1). Note that the control unit 100 also divides symbols such as “:” and “?” Into one phrase. Next, the control unit 100 extracts a plurality of phrase strings from the character string divided into phrases (step S1). The number of clauses extracted as a multiple clause string is limited to consecutive (connected) clauses, and the maximum number of clauses to be extracted is the same as the number of clauses included in the target character string.

Next, the control unit 100 calculates the appearance frequency of each phrase string (step S2). The control unit 100 records the calculated appearance frequency of each phrase string in the RAM 103. In addition, the control unit 100 sorts the phrase strings using the first character as a key for each phrase string such as a two-phrase string and a three-phrase string, and records them in the RAM 103.
As a result of the morphological analysis of each character string recorded as the input history, the control unit 100 sorts the phrase string using the first character as a key for each number of phrases included in the phrase string, and the appearance frequency of each phrase string. Etc. are recorded in the RAM 103.

  Next, when receiving input of a character string with the keyboard 116, the control unit 100 detects a phrase string to be excluded from the phrase string candidates displayed on the display 112 as input prediction candidates (step S3). Details of this processing will be described later with reference to the flowchart shown in FIG. This process is mainly executed by the third detection unit 203 and the frequency recalculation unit 204 shown in FIG.

  Next, the control unit 100 analyzes the meanings represented by the phrase strings, and groups the phrase strings having the same meaning into one group (step S4). Details of this processing will also be described later with reference to the flowchart shown in FIG. This process is mainly executed by the second detection unit 209 and the second selection unit 210 shown in FIG.

  Next, the control unit 100 creates a simplified notation for the phrase string, and groups the phrase strings having the same simplified notation into one group (step S5). This process will also be described later with reference to the flowchart shown in FIG. The shorthand notation is a notation in which only independent words are extracted from the phrase string. This process is mainly executed by the first detection unit 206 and the first selection unit 207 shown in FIG.

  Finally, the control unit 100 stores the recording information of the RAM 103 shown in FIG. 6 in the storage unit 104 and ends the process (step S6).

A procedure for detecting a phrase string to be excluded from the phrase string candidates displayed on the display 112 as input prediction candidates (details of the processing in step S3 shown in FIG. 7) will be described with reference to the flowchart shown in FIG.
First, the control unit 100 sets an initial value of a variable n for identifying a phrase string to n = 2 (step S11). Next, the control unit 100 selects a phrase string to be processed (hereinafter referred to as a phrase string to be processed) from the two phrase strings indicated by the variable n. Further, the control unit 100 sets the appearance frequency of the phrase string to be processed to the variable α, and sets the variable frec indicating the total appearance frequency to 0 (step S12). Note that the variable frec is a variable that counts the total appearance frequency of clause strings that are grouped (grouped) by the process described below. Further, the frequency of appearance of each phrase string has been calculated in step S2 shown in FIG.

Next, the control unit 100 sets m (m = n + 1) as a variable for identifying a phrase string other than the phrase string to be processed, which is a phrase string having a larger number of clauses than the phrase string to be processed (step n). S13). Note that m represents the number of clauses included in the clause string. For example, if the clause string to be processed is a two-phrase string (n = 2), the target phrase string is a three-phrase string having a larger number of phrases than the two-phrase string. In the following, the phrase string indicated by the variable m is referred to as an m phrase string.
The control unit 100 detects an m phrase string including the phrase string to be processed among the m phrase strings (step S14). For example, if the phrase string to be processed is “blood pressure:” (two phrase strings), the three phrase strings “blood pressure: normal” and “blood pressure:” include the phrase strings to be processed. When the m phrase string is detected (step S14 / YES), the control unit 100 adds the appearance frequency of the detected m phrase string to the variable frec for counting the total appearance frequency (step S15). Next, the control unit 100 determines whether there is another m phrase string that has not been determined in step S14 (step S16). If there is another m phrase string that has not been determined in step S14 (step S16 / YES), the control unit 100 repeats the processing from step S14. If it is determined that there is no other m phrase string that has not been determined in step S14 (step S16 / NO), the control unit 100 adds 1 to the value of m (step S17). And the control part 100 determines whether the value of m became larger than the phrase row | line | column with the largest number of phrases (step S18). If it is determined that the value of m is not greater than the phrase string with the maximum number of phrases (step S18 / NO), the control unit 100 repeats the process from step S14. If it is determined that the value of m is larger than the phrase string having the largest number of phrases (step S18 / YES), the control unit 100 determines whether or not the value of the variable α matches the value of the variable frec. Determination is made (step S19). If the control unit 100 determines that the values of the variable α and the variable frec match (step S19 / YES), the control unit 100 records a deletion flag in the record column of the phrase string to be processed in the RAM 103 (step S20). That is, the fact that the values of the variable α and the variable frec match indicates that the appearance frequency of the phrase string to be processed matches the appearance frequency of the detected m phrase string. Therefore, when the phrase string to be processed is input, it indicates that the user has always input the detected m phrase string. For example, in the example shown in FIGS. 5C and 5D, the three phrase strings including “blood pressure /:” in the two phrase strings are “blood pressure /: / abnormal” and “blood pressure /: / normal”. The total appearance frequency of the three-phrase string is 2, and the appearance frequency of “blood pressure /:” in the two-phrase string is also 2. Therefore, when “blood pressure:” is input, it indicates that the user inputs “blood pressure: abnormal” or “blood pressure: normal”. For this reason, the control unit 100 records a deletion flag indicating that it is excluded from the phrase string candidates to be displayed as the input prediction candidates in the recording column of the phrase string “blood pressure /:” in the RAM 103.

  Next, the control unit 100 determines whether there is another clause string having the same number of clauses as the number of clauses indicated by the variable n (step S21). For example, if the value of the variable n is 2, the control unit 100 determines whether there are other two phrase strings that are not selected as the phrase string to be processed. If it is determined that there is another phrase string in the variable n (step S21 / YES), the control unit 100 selects the other phrase string as a processing target phrase string, and repeats the processing from step S12 onward. If it is determined that there is no other phrase string (step S21 / NO), the control unit 100 adds 1 to the value of the variable n (step S22), and the value of the variable n is the phrase with the maximum number of phrases. It is determined whether or not the number of clauses in the column matches (step S23). If it is determined that they match (step S23 / YES), the control unit 100 ends this flow. If it is determined that they do not match (step S23 / NO), the control unit 100 repeats the processing from step S12.

Next, a process of analyzing the meanings of the phrase strings and grouping the phrase strings having the same meanings into one group (details of the process in step S4 shown in FIG. 7) will be described with reference to the flowchart shown in FIG.
The control unit 100 extracts an independent word from the phrase string for the phrase string other than the phrase string in which the deletion flag is recorded in the RAM 103 in the process of step S20 shown in FIG. 8 (step S31). Since the morpheme analysis is performed in step S1 shown in FIG. 7 and the result is recorded in the RAM 103, the control unit 100 refers to the result of the morpheme analysis recorded in the RAM 103 and extracts an independent word from the phrase string (step S1). S31).

  Next, the control unit 100 obtains a symbol representing the meaning of the extracted independent word (step S32). In the dictionary DB 402 of the storage unit 104, a dictionary in which the meaning of the independent word is recorded in advance is registered, and the control unit 100 refers to this dictionary to obtain a symbol representing the meaning of the independent word. Here, a case where English is used as a symbol representing the meaning of an independent word will be described as an example. For example, when the phrase string “open window” is targeted, the independent words are “window” and “open”, and the symbols representing the meanings of these independent words are “WINDOW” and “OPEN”. .

  Next, the control unit 100 sorts the obtained symbols in alphabetical order (step S33). In the example described above, since the symbols are “WINDOW” and “OPEN”, the sorting result is OPEN_WINDOW. Note that “_” is used in this specification as a symbol representing concatenation of words. The control unit 100 records the sorting result in the RAM 103.

  Next, the control unit 100 refers to the synonym dictionary stored in the storage unit 104, and acquires the headword of the independent word acquired in step S31 (step S34). In the synonym dictionary, a headword and a word synonymous with the headword are registered.

  Next, the control unit 100 selects a phrase string (hereinafter referred to as phrase string A) to be processed (step S35). Next, the control unit 100 detects a phrase string (hereinafter referred to as a phrase string B) having the same meaning as the phrase string A, or a matching headword of the independent word included in the phrase string A (hereinafter referred to as a phrase string B). S36). Since the symbols (English) representing the meanings of the independent words included in each phrase string are obtained in steps S32 and S33 and sorted in alphabetical order, the control unit 100 represents the meanings of the phrase string A and other phrase strings. The symbols are compared to find a phrase string that matches the symbols. Further, since the independent word headword included in each phrase string is acquired in step S34, the control unit 100 detects another phrase string in which the phrase string A matches the headword. The number of phrase strings B is not limited to one.

  When the phrase string B that has the same meaning as the phrase string A or matches the headword cannot be detected (step S37 / NO), the control unit 100 proceeds to the process of step S43. In step S43, the control unit 100 determines whether there is another phrase string that is not selected as the phrase string A among the phrase strings in which no deletion flag is recorded (step S43). If it is determined that there is a phrase string not selected as the phrase string A (step S43 / NO), the control unit 100 repeats the processing from step S35. If it is determined that there is no phrase string not selected as the phrase string A (step S43 / YES), the control unit 100 ends this flow.

When the phrase string B is detected in step S37 (step S37 / YES), the control unit 100 performs a process of grouping the phrase string B and the phrase string A together. In the grouping process, one of the phrase string A and the phrase string B is selected, and the selected character string is used as a phrase string representing the group. Further, a deletion flag is recorded in the recording column of the RAM 103 for the character string that has not been selected. Further, the appearance frequencies of the phrase string A and the phrase string B are added, and the added appearance frequency is recorded in the record column of the appearance frequency of the phrase string selected as the phrase string representing the group.
First, the control unit 100 compares the appearance frequencies of the phrase string A and the phrase string B (step S38). When it is determined that the appearance frequency of the detected phrase string B is higher than the appearance frequency of the phrase string A to be processed (step S38 / YES), the control unit 100 first records the phrase string C in the recording column of the RAM 103. A field is created, and the record information of the phrase string B is copied to the record field of the phrase string C. Further, the control unit 100 records a deletion flag in the record column of the phrase string A and the phrase string B in the RAM 103 (step S39). Then, the control unit 100 adds the appearance frequency of the phrase string A and the appearance frequency of the phrase string B, and records the addition result in the recording column of the RAM 103 as the appearance frequency of the phrase string C (step S40). If the frequency of occurrence of phrase string A is the same as the frequency of occurrence of phrase string B, or the frequency of occurrence of phrase string A is greater than the frequency of occurrence of phrase string B (step S38 / NO), the control unit 100 first creates a record column of the phrase string D in the record column of the RAM 103 and copies the record information of the phrase string A to the record column of the phrase string D. Further, the control unit 100 records a deletion flag in the record column of the phrase string A and the phrase string B in the RAM 103 (step S41). Then, the control unit 100 adds the appearance frequency of the phrase string A and the appearance frequency of the phrase string B, and records the addition result in the recording column of the RAM 103 as the appearance frequency of the phrase string D (step S42). Then, the control part 100 transfers to the process of step S43 mentioned above.
When the frequency of occurrence of the phrase string A and the phrase string B is the same, the longer phrase string, that is, the greater number of included phrases is left as an input prediction candidate, and the number of phrases is smaller. A deletion flag may be recorded on the other side.

Next, referring to the flowchart shown in FIG. 10, a description will be given of a process of creating a simplified notation of a phrase string, grouping together phrase strings having the same simplified notation, and grouping them (the details of the process in step S <b> 5 shown in FIG. 7). To do.
First, the control unit 100 refers to the result of the morphological analysis recorded in the RAM 103, and records the deletion flag in the record column of the phrase string other than the particle where the phrase string ends with a comma and immediately before the comma. For example, the deletion flag is recorded in the record column of the phrase string “blood pressure was normal” (step S51).
Next, the control unit 100 refers to the result of the morphological analysis recorded in the RAM 103, and extracts an independent word from the phrase string for the phrase string other than the phrase string in which the deletion flag is recorded (step S52). Next, the control unit 100 creates a simplified notation R created by connecting only the extracted independent words. For example, the simplified notation R of “open window” is “open window”, and the simplified notation R of “blood pressure: normal” is “normal blood pressure”.

  Next, the control unit 100 selects a phrase string (hereinafter referred to as phrase string E) to be processed (step S53). Next, the control unit 100 detects a phrase string (hereinafter referred to as a phrase string F) in which the phrase string E to be processed and the simplified notation R match (step S54). The number of phrase strings F is not limited to one. When a phrase string F having the same simplified notation as the phrase string E is detected (step S55 / YES), the control unit 100 compares the appearance frequencies of the phrase string E and the phrase string F. The control unit 100 selects one of the phrase string E and the phrase string F that has the highest appearance frequency, and groups the other phrase string together into the selected phrase string. That is, the appearance frequency of character strings is compared, and the phrase string with the lower appearance frequency is excluded from the input prediction candidates. In addition, the appearance frequency of the phrase string having the lower appearance frequency is added to the phrase string having the higher appearance frequency, and the addition result is set as the appearance frequency of the phrase string having the higher appearance frequency. When it is determined that the appearance frequency of the phrase string F is higher than the appearance frequency of the phrase string E (step S56 / YES), the control unit 100 first creates a recording column for the phrase string G in the recording column of the RAM 103, Copy the record information of the phrase string F to the record column of the string G. Further, the control unit 100 records a deletion flag in the record column of the phrase string E and the phrase string F in the RAM 103 (step S57). Then, the control unit 100 adds the appearance frequency of the phrase string E and the appearance frequency of the phrase string F, and records the addition result in the RAM 103 as the appearance frequency of the phrase string G (step S58). Further, when the appearance frequency of the phrase string E and the appearance frequency of the phrase string F are the same, or when the appearance frequency of the phrase string E is higher than the appearance frequency of the phrase string F (step S56 / NO), the control unit 100 first creates a record column of the phrase string H in the record column of the RAM 103 and copies the record information of the phrase string E to the record column of the phrase string H. Further, the control unit 100 records a deletion flag in the record column of the phrase string E and the phrase string F in the RAM 103 (step S59). Then, the control unit 100 adds the appearance frequency of the phrase string E and the appearance frequency of the phrase string F, and records the addition result in the RAM 103 as the appearance frequency of the phrase string H (step S60). Thereafter, the control unit 100 proceeds to the process of step S61. In step S61, the control unit 100 determines whether there is a phrase string that is not selected as the phrase string E among the phrase strings in which no deletion flag is recorded (step S61). If it is determined that there is a phrase string not selected as the phrase string E among the phrase strings for which no deletion flag is recorded (step S61 / NO), the control unit 100 repeats the processing from step S53. If it is determined that there is no phrase string not selected as the phrase string E (step S61 / YES), the control unit 100 ends this flow.

  In step S55, if the phrase string F in which the simplified notation of the phrase string E matches the simplified notation cannot be detected (step S55 / NO), the control unit 100 proceeds to the process of step S62. To do. In step S62, the control unit 100 determines whether there is a phrase string (hereinafter referred to as phrase string I) in which the simplified notation and the simplified notation of the phrase string E coincide with each other. That is, the control unit 100 determines whether or not there is a phrase string including the simplified notation of the phrase string E. When the control unit 100 detects a phrase string I whose simplified notation matches forward (step S63 / YES), the control unit 100 first creates a record column for the phrase string J in the record field of the RAM 103, and then sets the phrase string in the record column for the phrase string J. The recording information of I is copied (step S64). Further, the control unit 100 records a deletion flag in the record column of the phrase string E and the phrase string I in the RAM 103 (step S64). Then, the control unit 100 adds the appearance frequency of the phrase string E and the appearance frequency of the phrase string I, and records the addition result in the RAM 103 as the appearance frequency of the phrase string J (step S65). Then, the control part 100 transfers to the process of step S61 mentioned above.

The control unit 100 stores the information recorded in the RAM 103 through the above-described procedure in the storage unit 104 as an input prediction candidate DB. In addition, when the control unit 100 receives a character string with the keyboard 116, the control unit 100 refers to the input prediction candidate DB stored in the storage unit 104 and causes the display 112 to display the input prediction candidate for the received character string.
The phrase string recorded as the input prediction candidate in the input prediction candidate DB is obtained by grouping phrase strings having the same meaning and selecting and registering one phrase string in the group. The frequency of the selected phrase string is the total usage frequency of each phrase string included in the group. Therefore, the frequency of use can be aggregated in units of groups of character strings having the same meaning, and the character strings included in the group of meanings having a high usage frequency can be ranked higher in the input prediction candidates.
The phrase string created from the character string is generated by extracting a plurality of phrases from one sentence. In other words, the phrase string displayed as the input prediction candidate is not based on the character string that has been input from when the enter key of the keyboard 116 is pressed until it is pressed next. Therefore, it is possible to take out an effective phrase string having a high frequency from the one sentence in units of one sentence input by the user.

  In the above description, the deletion flag is recorded in the phrase string excluded from the input prediction candidates and is not displayed as the input prediction candidates. However, when the number of candidates to be displayed as input prediction candidates is small, the input prediction candidates may be searched from the phrase string including the phrase string in which the deletion flag is recorded.

Next, a display example of input prediction candidates displayed on the display 112 will be described.
In the first display example shown in FIG. 11, a main window 501 and a sub window 502 are displayed side by side. The main window 501 displays conversion candidates obtained by converting the character string input by the user (“Ketsuatsu” in the example shown in FIG. 11) by the kana-kanji conversion processing unit 250. In addition, when one of the conversion candidates (for example, “blood pressure”) displayed on the main window 501 is selected by the mouse 117 or the like, the input prediction candidate of the selected conversion candidate is displayed on the sub window 502. For example, when the conversion candidate to be displayed on the main window 501 in the process of the kana-kanji conversion processing unit 250 is “blood pressure”, the input prediction candidates “blood pressure abnormality” and “blood pressure: normal” are displayed on the sub window 502. These input prediction candidates are phrase strings including the characters “blood pressure” and having the highest appearance frequency.

  In the second display example shown in FIG. 12, the main window 501 is the only window to be displayed, and conversion candidates and input prediction candidates are displayed in the main window 501. In the example shown in FIG. 12, only one conversion candidate converted by the kana-kanji conversion processing unit 250 is displayed, and the input prediction candidate of this conversion candidate is displayed below the conversion candidate.

  In the third display example shown in FIG. 13, the display of the main window 501 is the same as that of the first display example shown in FIG. 11, but in the sub-window 502, a narrowed word for narrowing down input prediction candidates to be displayed. A display 503 requesting the input of is displayed. When a narrowed word is input to the sub-window 502, the input prediction candidate search unit 300 searches the input prediction candidate DB using the input narrowed word as a search key. The input prediction candidate search unit 300 takes out the phrase having the highest appearance frequency (the top two in the example shown in FIG. 13) from the phrase string including the conversion candidate displayed in the main window 501 and the narrowed-down word, and outputs the sub-window 502 is displayed.

In the fourth display example shown in FIG. 14A, the display of the main window 501 is the same as the first display example shown in FIG. In addition, when the conversion candidate displayed on the main window 501 is input to the sub-window 502, the word candidates input following the conversion candidate are displayed. The input prediction candidate search unit 300 searches the input prediction candidate DB, and when a conversion candidate word is input, searches for a phrase string including this word. The input prediction candidate search unit 300 extracts independent words included in the detected phrase string and displays the extracted independent words on the subwindow 502 as narrowed words. Further, the input prediction candidate search unit 300 displays in the sub-window 502 how many phrase strings including the conversion candidate word and the narrowed word.
When one of the narrowed words displayed on the sub window 502 is selected by operating the mouse 117 or the like, a phrase string including the selected narrowed word and the conversion candidate word is displayed. In the window for displaying the phrase string, a space may be provided between the sub-windows 502 as shown in FIG. 14B, and a window 504 for displaying the phrase string may be displayed in this space. As shown in FIG. 14C, a window 504 for displaying a phrase string may be displayed on the side of the sub window 502 that does not overlap with the main window 501.

  Further, the input prediction candidates to be displayed in the main window 501 and the sub window 502 may be displayed by changing the character font and background color according to the appearance frequency. For example, input prediction candidates whose appearance frequency is 10 or more are displayed using a bold font, and the background color in the subwindow is highlighted in red. An input prediction candidate whose appearance frequency is 2 or less is displayed using a gray font.

  The embodiment described above is a preferred embodiment of the present invention. However, the present invention is not limited to this, and various modifications can be made without departing from the scope of the present invention.

[Appendix 1]
A plurality of partial characters including at least two or more words or clauses included in each character string data by analyzing the plurality of character string data received by the input receiving means and analyzing the plurality of character string data stored in the storage means Conversion means for converting to column data;
Frequency counting means for counting the appearance frequency of the partial character string data converted by the conversion means in the plurality of character string data;
From the partial character string data converted by the conversion means, an independent word extraction means for extracting an independent word;
The partial character string data to be processed is selected from the partial character string data converted by the conversion means, and the independent character included in the partial character string data to be processed is referred to by referring to the extraction result of the independent word extraction means. First detection means for detecting other partial character string data including a word as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected by the first detection means and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted by the frequency counting means First selection means for selecting character string data, and storing the selected partial character string data in the storage means as input prediction candidates for supporting input of the character string;
Calculate a total value of appearance frequencies of the partial character string data to be processed and the other partial character string data detected by the first detection unit, and use the calculated total value as the use frequency of the input prediction candidate A frequency calculation means for storing in the storage means;
When input of character string data is received by the input receiving means, based on the use frequency stored in the storage means, partial character string data including the character string data that has been accepted is selected, and the selected partial character Selection means for displaying column data on the display means;
A character input support device comprising:
[Appendix 2]
Dictionary storage means for storing a dictionary in which synonyms indicating the same meaning are registered;
Referring to the dictionary storage means, synonym extraction means for extracting synonyms of the independent words extracted by the independent word extraction means;
The partial character string data to be processed is selected from the partial character string data converted by the conversion means, and other partial character string data including synonyms of independent words included in the selected partial character string data, Second detection means for detecting as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected by the second detection means and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted by the frequency counting means Second selection means for selecting character string data and storing it as the input prediction candidates in the storage means,
The frequency calculation means calculates a total value of appearance frequencies of the partial character string data to be processed and the other partial character string data detected by the second detection means, and the calculated total value is calculated as the input prediction. The character input support device according to appendix 1, wherein the storage means stores the candidate usage frequency.
[Appendix 3]
A part that includes partial character string data to be processed from the partial character string data converted by the conversion means and includes more words or phrases than the number of words or phrases included in the selected partial character string data Detecting other partial character string data including word or phrase included in the processing target partial character string data and storing the detected other partial character string data as the input prediction candidate A third detecting means for storing in the means;
The frequency calculating means calculates a total value of appearance frequencies of the other partial character string data detected by the third detecting means and the partial character string data to be processed, and the calculated total value is calculated as the input prediction. The character input support device according to appendix 1 or 2, characterized in that the storage means stores the candidate use frequency.
[Appendix 4]
A program for operating a computer as a character input support system for supporting character input,
The computer,
A plurality of partial characters including at least two or more words or clauses included in each character string data by analyzing the plurality of character string data received by the input receiving means and analyzing the plurality of character string data stored in the storage means Conversion means for converting to column data;
Frequency counting means for counting the appearance frequency of the partial character string data converted by the conversion means in the plurality of character string data;
From the partial character string data converted by the conversion means, an independent word extraction means for extracting an independent word;
The partial character string data to be processed is selected from the partial character string data converted by the conversion means, and the independent character included in the partial character string data to be processed is referred to by referring to the extraction result of the independent word extraction means. First detection means for detecting other partial character string data including a word as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected by the first detection means and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted by the frequency counting means First selection means for selecting character string data, and storing the selected partial character string data in the storage means as input prediction candidates for supporting input of the character string;
Calculate a total value of appearance frequencies of the partial character string data to be processed and the other partial character string data detected by the first detection unit, and use the calculated total value as the use frequency of the input prediction candidate A frequency calculation means for storing in the storage means;
When input of character string data is received by the input receiving means, based on the use frequency stored in the storage means, partial character string data including the character string data that has been accepted is selected, and the selected partial character A program that functions as selection means for displaying column data on a display means.
[Appendix 5]
The computer,
A synonym extracting means for extracting a synonym of the independent word extracted by the independent word extracting means with reference to a dictionary in which synonyms indicating the same meaning are registered;
The partial character string data to be processed is selected from the partial character string data converted by the conversion means, and other partial character string data including synonyms of independent words included in the selected partial character string data, Second detection means for detecting as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected by the second detection means and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted by the frequency counting means Select character string data and function as second selection means for storing the input prediction candidates in the storage means,
The frequency calculation means calculates a total value of appearance frequencies of the partial character string data to be processed and the other partial character string data detected by the second detection means, and the calculated total value is calculated as the input prediction. The program according to appendix 4, wherein the storage unit stores the candidate usage frequency.
[Appendix 6]
The computer,
A part that includes partial character string data to be processed from the partial character string data converted by the conversion means and includes more words or phrases than the number of words or phrases included in the selected partial character string data Detecting other partial character string data including word or phrase included in the processing target partial character string data and storing the detected other partial character string data as the input prediction candidate Function as a third detection means to be stored in the means,
The frequency calculating means calculates a total value of appearance frequencies of the other partial character string data detected by the third detecting means and the partial character string data to be processed, and the calculated total value is calculated as the input prediction. The program according to appendix 4 or 5, which is stored in the storage means as a candidate use frequency.
[Appendix 7]
A character input support method executed by a computer device that supports character input,
A plurality of partial characters including at least two or more words or clauses included in each character string data by analyzing the plurality of character string data received by the input receiving means and analyzing the plurality of character string data stored in the storage means A conversion step for converting to column data;
A frequency counting step of counting the appearance frequency of the partial character string data converted in the conversion step in the plurality of character string data;
An independent word extraction step of extracting an independent word from the partial character string data converted by the conversion step,
The partial character string data to be processed is selected from the partial character string data converted by the conversion step, and the independent character included in the partial character string data to be processed is referred to by referring to the extraction result of the independent word extraction step. A first detection step of detecting other partial character string data including a word as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected in the first detection step and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted in the frequency counting step A first selection step of selecting character string data and storing the selected partial character string data in the storage means as an input prediction candidate that supports input of the character string;
Calculate the total value of the appearance frequency of the partial character string data to be processed and the other partial character string data detected in the first detection step, and use the calculated total value as the use frequency of the input prediction candidate A frequency calculation step of storing in the storage means;
When input of character string data is received by the input receiving means, based on the use frequency stored in the storage means, partial character string data including the character string data that has been accepted is selected, and the selected partial character A character input support method, wherein the computer device executes a selection step of displaying column data on a display means.
[Appendix 8]
The computer device includes:
A synonym extraction step of extracting a synonym of the independent word extracted in the independent word extraction step with reference to a dictionary in which synonyms indicating the same meaning are registered;
The partial character string data to be processed is selected from the partial character string data converted in the conversion step, and the other partial character string data including synonyms of independent words included in the selected partial character string data, A second detection step of detecting as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected in the second detection step and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted in the frequency counting step A second selection step of selecting character string data and causing the storage means to store it as the input prediction candidate,
The frequency calculation step calculates a total value of appearance frequencies of the partial character string data to be processed and the other partial character string data detected in the second detection step, and calculates the calculated total value as the input prediction 8. The character input support method according to appendix 7, wherein the storage means stores the candidate usage frequency.
[Appendix 9]
The computer device includes:
A portion that includes partial character string data to be processed from the partial character string data converted in the conversion step and includes more words or phrases than the number of words or phrases included in the selected partial character string data Detecting other partial character string data including word or phrase included in the processing target partial character string data and storing the detected other partial character string data as the input prediction candidate Further executing a third detection step to be stored in the means;
The frequency calculation step calculates a total value of appearance frequencies of the other partial character string data detected in the third detection step and the partial character string data to be processed, and calculates the calculated total value as the input prediction The character input support method according to appendix 7 or 8, wherein the storage means stores the candidate use frequency.

DESCRIPTION OF SYMBOLS 1 Computer apparatus 100 Control part 104 Storage part 112 Display 115 Operation part 116 Keyboard 117 Mouse 201 Conversion part 202 Frequency counting part 203 3rd detection part 204 Frequency recalculation part 205 Autonomous word extraction part 206 1st detection part 207 1st selection part 208 synonym extraction unit 209 second detection unit 210 second selection unit

Claims (5)

A plurality of partial characters including at least two or more words or clauses included in each character string data by analyzing the plurality of character string data received by the input receiving means and analyzing the plurality of character string data stored in the storage means Conversion means for converting to column data;
Frequency counting means for counting the appearance frequency of the partial character string data converted by the conversion means in the plurality of character string data;
From the partial character string data converted by the conversion means, an independent word extraction means for extracting an independent word;
The partial character string data to be processed is selected from the partial character string data converted by the conversion means, and the independent character included in the partial character string data to be processed is referred to by referring to the extraction result of the independent word extraction means. First detection means for detecting other partial character string data including a word as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected by the first detection means and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted by the frequency counting means First selection means for selecting character string data, and storing the selected partial character string data in the storage means as input prediction candidates for supporting input of the character string;
Calculate a total value of appearance frequencies of the partial character string data to be processed and the other partial character string data detected by the first detection unit, and use the calculated total value as the use frequency of the input prediction candidate A frequency calculation means for storing in the storage means;
When input of character string data is received by the input receiving means, based on the use frequency stored in the storage means, partial character string data including the character string data that has been accepted is selected, and the selected partial character Selection means for displaying column data on the display means;
A character input support device comprising: Dictionary storage means for storing a dictionary in which synonyms indicating the same meaning are registered;
Referring to the dictionary storage means, synonym extraction means for extracting synonyms of the independent words extracted by the independent word extraction means;
The partial character string data to be processed is selected from the partial character string data converted by the conversion means, and other partial character string data including synonyms of independent words included in the selected partial character string data, Second detection means for detecting as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected by the second detection means and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted by the frequency counting means Second selection means for selecting character string data and storing it as the input prediction candidates in the storage means,
The frequency calculation means calculates a total value of appearance frequencies of the partial character string data to be processed and the other partial character string data detected by the second detection means, and the calculated total value is calculated as the input prediction. The character input support apparatus according to claim 1, wherein the storage unit stores the candidate usage frequency. A part that includes partial character string data to be processed from the partial character string data converted by the conversion means and includes more words or phrases than the number of words or phrases included in the selected partial character string data Detecting other partial character string data including word or phrase included in the processing target partial character string data and storing the detected other partial character string data as the input prediction candidate A third detecting means for storing in the means;
The frequency calculating means calculates a total value of appearance frequencies of the other partial character string data detected by the third detecting means and the partial character string data to be processed, and the calculated total value is calculated as the input prediction. The character input support apparatus according to claim 1, wherein the storage unit stores the candidate usage frequency. A program for operating a computer as a character input support system for supporting character input,
The computer,
A plurality of partial characters including at least two or more words or clauses included in each character string data by analyzing the plurality of character string data received by the input receiving means and analyzing the plurality of character string data stored in the storage means Conversion means for converting to column data;
Frequency counting means for counting the appearance frequency of the partial character string data converted by the conversion means in the plurality of character string data;
From the partial character string data converted by the conversion means, an independent word extraction means for extracting an independent word;
The partial character string data to be processed is selected from the partial character string data converted by the conversion means, and the independent character included in the partial character string data to be processed is referred to by referring to the extraction result of the independent word extraction means. First detection means for detecting other partial character string data including a word as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected by the first detection means and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted by the frequency counting means First selection means for selecting character string data, and storing the selected partial character string data in the storage means as input prediction candidates for supporting input of the character string;
Calculate a total value of appearance frequencies of the partial character string data to be processed and the other partial character string data detected by the first detection unit, and use the calculated total value as the use frequency of the input prediction candidate A frequency calculation means for storing in the storage means;
When input of character string data is received by the input receiving means, based on the use frequency stored in the storage means, partial character string data including the character string data that has been accepted is selected, and the selected partial character A program that functions as selection means for displaying column data on a display means. A character input support method executed by a computer device that supports character input,
A plurality of partial characters including at least two or more words or clauses included in each character string data by analyzing the plurality of character string data received by the input receiving means and analyzing the plurality of character string data stored in the storage means A conversion step for converting to column data;
A frequency counting step of counting the appearance frequency of the partial character string data converted in the conversion step in the plurality of character string data;
An independent word extraction step of extracting an independent word from the partial character string data converted by the conversion step,
The partial character string data to be processed is selected from the partial character string data converted by the conversion step, and the independent character included in the partial character string data to be processed is referred to by referring to the extraction result of the independent word extraction step. A first detection step of detecting other partial character string data including a word as partial character string data having the same meaning as the partial character string data to be processed;
Of the other partial character string data detected in the first detection step and the partial character string data to be processed, a part according to the length of the partial character string data or the appearance frequency counted in the frequency counting step A first selection step of selecting character string data and storing the selected partial character string data in the storage means as an input prediction candidate that supports input of the character string;
Calculate the total value of the appearance frequency of the partial character string data to be processed and the other partial character string data detected in the first detection step, and use the calculated total value as the use frequency of the input prediction candidate A frequency calculation step of storing in the storage means;
When input of character string data is received by the input receiving means, based on the use frequency stored in the storage means, partial character string data including the character string data that has been accepted is selected, and the selected partial character A character input support method, wherein the computer device executes a selection step of displaying column data on a display means.

Images