Input assistance program, input assistance method, and input assistance apparatus

Abstract

To improve the accuracy of detecting an input error made by a user.SOLUTION: An input assistance apparatus 100 specifies, from an input character string 101, a portion 102 of which a value indicating unnatural character sequence is smaller than a first reference value and equal to or larger than a second reference value which is smaller than the first reference value. The input assistance apparatus 100 applies predetermined correction processing to the specified portion 102 to create reading candidates 104 of the portion 102. The input assistance apparatus 100 refers to a storage unit 110 which stores contextual relationships between semantically related words which are input in a row, to determine whether the created reading candidates 104 include a reading candidate which is semantically related to at least a word (word 103) before or after the portion 102 in the input character string 101. The input assistance apparatus 100 detects an input error in the portion 102 when determining that there is a semantically related reading candidate.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an input support program, an input support method, and an input support device for assisting character input performed based on a user's operation.

Background Art

[0002] Conventionally, in a Japanese input system, there is a technique for automatically correcting input errors made by a user. For example, when detecting a vowel omission (input error) from an input character string, a vowel is supplemented at the omitted position to estimate the reading, and a result obtained by converting the estimated reading into a Chinese character or the like is output as a conversion candidate.

[0003] As a prior art for assisting character input, there is one in which an input reading string temporary storage unit stores an input reading string, an extended candidate generation unit generates an extended candidate by interpreting a partial reading string of the reading string in the input reading string temporary storage unit as another reading string, and a candidate limitation unit selects a subset of the extended candidates generated by the extended candidate generation unit according to the selected operation method.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in the prior art, there are cases where it is impossible to detect input errors made by a user from an input character string. For example, if an input error cannot be detected, the portion corresponding to the input error is not repaired, and a conversion result that the user does not desire is output.

[0006] On one side, an object of the present invention is to provide an input support program, an input support method, and an input support device that improve the detection accuracy of input errors by a user.

Means for Solving the Problem

[0007] In order to solve the above-described problems and achieve the object, the input support program according to the present invention, based on the arrangement of characters in the input string, identifies a portion where a value indicating unnaturalness of the character arrangement in the string is smaller than a first reference value for detecting an input error and is equal to or greater than a second reference value smaller than the first reference value, creates a reading candidate for the portion by performing a predetermined correction process on the identified portion, refers to a storage unit that stores the context between words having a semantic connection that are continuously input, and determines whether there is a reading candidate among the created reading candidates that has a semantic connection with at least one of the words before and after the portion in the string. When there is a reading candidate having a semantic connection with any of the words, the computer is caused to execute a process of detecting an input error at the portion.

[0008] Further, the input support program according to the present invention, in the above invention, when an input error at the portion is detected, causes the computer to execute a process of outputting a conversion candidate corresponding to the string based on the created reading candidate.

[0009] Further, the input support program according to the present invention, in the above invention, the output process outputs a conversion candidate corresponding to the string based on the created reading candidate and the original reading according to the arrangement of characters at the portion.

[0010] In addition, in the above invention, the input support program according to the present invention causes the computer to execute a process of creating a first conversion candidate corresponding to the character string based on the created reading candidate, and creating a second conversion candidate corresponding to the character string based on the original reading, and the output process preferentially outputs a conversion candidate corresponding to a reading candidate having a semantic connection with any one of the created first conversion candidate and the second conversion candidate.

[0011] In addition, in the above invention, in the input support program according to the present invention, the storage unit further stores the strength of the semantic connection between the words, and the detection process detects an input error at the location when there is a reading candidate having a semantic connection and the strength of the semantic connection between the reading candidate having a semantic connection and any one of the words is equal to or greater than a threshold value.

[0012] In addition, the input support method according to the present invention specifies a location where a value indicating unnaturalness of the character sequence from the input character string is smaller than a first reference value for detecting an input error and is equal to or greater than a second reference value smaller than the first reference value, creates a reading candidate for the location by performing a predetermined correction process on the specified location, refers to a storage unit that stores the context between continuously input words having a semantic connection, and determines whether there is a reading candidate having a semantic connection with at least any one of the words before and after the location in the character string among the created reading candidates. When there is a reading candidate having a semantic connection with any one of the words, the computer executes a process of detecting an input error at the location.

[0013] In addition, the input support device according to the present invention, based on the order of characters in the input string, identifies a location where a value indicating the unnaturalness of the character order in the string is less than a first reference value for detecting an input error and is equal to or greater than a second reference value smaller than the first reference value, creates a reading candidate for the identified location by performing a predetermined correction process on the identified location, refers to a storage unit that stores the context between words having a semantic connection that are continuously input, determines whether there is a reading candidate among the created reading candidates that has a semantic connection with at least one of the words before and after the location in the string, and when there is a reading candidate having a semantic connection with any of the words, detects an input error at the location, and has a control unit that executes the process.

Advantages of the Invention

[0014] According to one aspect of the present invention, there is an effect that the detection accuracy of input errors by the user can be improved.

Brief Description of the Drawings

[0015]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

[0016] (Embodiment) Hereinafter, embodiments of an input support program, an input support method, and an input support device according to the present invention will be described in detail with reference to the drawings.

[0017] FIG. 1 is an explanatory diagram showing an example of an embodiment of the input support method according to the embodiment. In FIG. 1, the input support device 100 is a computer that supports character input performed based on a user's operation. The input support device 100 is, for example, a PC (Personal Computer). The input support device 100 may be a tablet PC, a smartphone, or the like. Further, the input support device 100 may be a server that can be connected from a PC or the like used by the user.

[0018] The user's operation is performed using an input device such as a keyboard or a touch panel, for example. In the case of Japanese input, there are input modes such as Roman character input and kana input. In Roman character input, when inputting Japanese, for example, Roman characters combining the consonants and vowels of the characters are input. Kana input is input using kana characters written on a keyboard or the like.

[0019] In character input performed based on the user's operation, input errors may occur. Examples of input errors include character omission, input of extra characters, and incorrect input order of characters. If the input error made by the user can be automatically corrected, it is convenient because the user can save the trouble of re-entering.

[0020] As a method for correcting input errors, for example, when a vowel is missing, a vowel is added to the missing part. In addition, a machine learning method can be used to learn the natural order of romanized characters in Japanese, and when an input error occurs, characters can be added, unnecessary characters can be deleted, or the order of characters can be changed.

[0021] However, in the conventional technology, there are cases where input errors made by a user cannot be detected. If an input error cannot be detected, the part corresponding to the input error will not be corrected, and an undesired conversion result will be output.

[0022] For example, there is a conventional technique that detects as an input error any part of an input character string where the value indicating the unnaturalness of the character sequence is equal to or exceeds a predetermined reference value. Suppose a user tries to input the word "ichikenrakuchaku" (completely settled) and inputs the character string "ikenrakuchaku." The character "ikken" should have been input, but an input error occurred in which the letter "k" was omitted.

[0023] However, "iken" is a word that actually exists in the dictionary, and although the value indicating the unnaturalness of the character arrangement is somewhat high, it is not high enough to detect an input error. In this case, the conventional technology cannot detect "iken" as an input error, and the part corresponding to the input error is not corrected, resulting in the output of a conversion result that the user does not want (for example, "opinion settlement").

[0024] On the other hand, it is possible to lower the standard value for detecting input errors so that even parts where the character sequence is somewhat unnatural are detected as input errors. However, if the standard value is lowered too much, there is a problem that input errors will be detected even in parts that are correctly entered, and as a result, conversion results that the user does not want will be output.

[0025] Therefore, in this embodiment, when detecting input errors from the input string, for parts where input errors are suspected, an input support method that improves the detection accuracy of input errors by the user using the semantic connection with the preceding and following words will be described. Here, a processing example of the input support device 100 will be described. The processing example of the input support device 100 is executed, for example, in kana-kanji conversion processing. The kana-kanji conversion processing is a process of converting the input string into a string including kanji or a sentence including kanji.

[0026] (1) Based on the character sequence of the input string 101, the input support device 100 identifies, from the input string 101, a location where the value indicating the unnaturalness of the character sequence is smaller than the first reference value and is equal to or greater than a second reference value smaller than the first reference value. Here, the input string 101 is a string input based on the user's operation.

[0027] The input string 101 is a string before confirmation. For example, in the case of Japanese input in the Roman character input mode, the input string 101 corresponds to the input Roman character spelling or the result of converting the Roman character spelling into kana. Also, in the case of Japanese input in the kana input mode, the input string 101 corresponds to the input kana spelling (keystrokes).

[0028] Input errors include, for example, character omission, input of extra characters, and incorrect input order of characters. The value indicating the unnaturalness of the character sequence is an index value for evaluating the character sequence, and the higher the value, the more unnatural the character sequence. The value indicating the unnaturalness of the character sequence is calculated, for example, based on the plausibility of the transition between adjacent characters.

[0029] For example, in the case of Japanese input in the Roman character input mode, the plausibility of the transition between adjacent characters can be obtained by learning the natural sequence of Roman characters as Japanese. The value indicating the unnaturalness of the character sequence is determined, for example, in units of clauses or words included in the input string 101.

[0030] The first reference value can be set arbitrarily, for example, to a value that can be considered to deviate from the natural arrangement of characters. The second reference value is a value smaller than the first reference value, for example, to a value that can be considered to be a slightly unnatural arrangement of characters, but not enough to be detected as an input error. This allows the location (for example, a phrase or word) where an input error is suspected to be present to be identified.

[0031] In the example of FIG. 1, it is assumed that Japanese is input in the romaji input mode, and the input character string 101 is "ikennrakuchaku". "ikennrakuchaku" is the input romaji spelling. "ikennrakuchaku" is "ikennrakuchaku" converted into kana.

[0032] Here, it is assumed that the user is attempting to input the word "ichiken-rakuchaku" (all clear), and inputs character string 101. It is also assumed that a value indicating the unnaturalness of the character arrangement of portion 102 corresponding to "ikenn" in input character string 101 is smaller than a first reference value and equal to or greater than a second reference value that is smaller than the first reference value. In this case, portion 102 corresponding to "ikenn" is identified from input character string 101. Portion 102 corresponding to "ikenn" corresponds to a portion suspected of being an input error.

[0033] (2) The input support device 100 performs a predetermined correction process on the identified portion 102 to generate a reading candidate for the portion 102. Here, the predetermined correction process is a process of adding characters, deleting unnecessary characters, or changing the order of characters in the portion to be processed (e.g., the portion 102). The reading candidate corresponds to a repair candidate obtained by repairing the sequence of romaji in the portion 102 and converting it into kana, for example.

[0034] In the example of FIG. 1, for the location 102 corresponding to "ikenn", the input support device 100, for example, determines that the character "k" is missing between "i" and "k" from the natural order of the learned Roman characters in Japanese, and performs a correction process to supplement the character "k". As a result, a reading candidate 104 of "ikkenn" is created.

[0035] (3) The input support device 100 refers to the storage unit 110 and determines whether there is a reading candidate among the created reading candidates 104 that has a semantic connection with at least one of the words before and after the location 102 in the input character string 101. Here, the storage unit 110 stores the sequential relationship between words with a semantic connection that are continuously input.

[0036] Words that are continuously input are, for example, words that are continuously input. Also, words that are continuously input may be words that are input with a particle or auxiliary verb such as "te ni o wa" in between. A semantic connection is a relationship in which two or more words are combined to form a new meaningful word.

[0037] For example, a semantic connection may be a relationship in which two or more words are combined to form a compound word. Also, a semantic connection may be a relationship in which two or more words are combined to form an idiomatic expression. Also, a semantic connection may be a relationship between a modifier and a modified word.

[0038] In the example of FIG. 1, it is assumed that the storage unit 110 stores the sequential relationship (word order) between the word "ikkenn" and the word "rakuchaku" as the sequential relationship between words with a semantic connection. The word "ikkenn" and the word "rakuchaku" form a compound word "ikkenn rakuchaku". The sequential relationship between the words is that the word "ikkenn" is in the front and the word "rakuchaku" is in the back.

[0039] Specifically, for example, the input support device 100 extracts at least one of the words before and after the identified location 102 from the input character string 101. The previous word is, for example, the word immediately before the location 102 in the input character string 101. The following word is, for example, the word immediately after the location 102 in the input character string 101. However, particles and auxiliary verbs such as "てにをは" may be excluded from the extraction target.

[0040] In the example of FIG. 1, the word 103 "らくちゃく" immediately after the location 102 is extracted from the input character string 101. The word 103 matches the word "らくちゃく" stored in the storage unit 110. The word "らくちゃく" has a semantic connection with the word "いっけん" and appears after the word "いっけん". In this case, the input support device 100 refers to the storage unit 110 and determines that there is a reading candidate in the created reading candidates 104 that appears before the extracted word 103 and has a semantic connection with the extracted word 103.

[0041] (4) When there is a reading candidate that has a semantic connection with the extracted word 103, the input support device 100 detects an input error at the location 102. Although the arrangement of characters at the location 102 is somewhat unnatural, it cannot be simply determined as an input error based only on the value indicating the unnaturalness of the character arrangement.

[0042] Regarding the location 102, if there is a semantic connection between the created reading candidate 104 and the following word 103, it can be said that the reading candidate 104 may be the correct reading. Therefore, for example, if there is a semantic connection between the created reading candidate 104 and the following word 103 for that location 102, the input support device 100 determines that it is not the arrangement of characters originally intended to be input and detects an input error at the location 102. This process corresponds to, for example, lowering the reference value used when detecting an input error based on the value indicating the unnaturalness of the character arrangement from the first reference value to the second reference value.

[0043] Thus, according to the input support device 100, when detecting input errors from the input string, regarding the locations where input errors are suspected, the semantic connection with the preceding and succeeding words can be utilized to improve the detection accuracy of input errors by the user. For example, even if the value indicating the unnaturalness of the character arrangement is not so high as to deviate from the natural arrangement of characters, the input support device 100 can determine that an input error has occurred based on the semantic connection with the preceding and succeeding words.

[0044] In the example of FIG. 1, when the user attempts to input the word "ippan rakuchaku", for the location 102 where "ikenn" was mistakenly input, an input error can be detected from the semantic connection between the created reading candidate 104 and the subsequent word 103. Thereby, the input support device 100 can repair the location 102 where the input error occurred in the input string 101, and prevent the output of conversion results that the user does not desire. Specifically, for example, the input support device 100 can output the conversion candidate "ippan rakuchaku" desired by the user by repairing the location 102 in the input string 101 to "ikkenn".

[0045] (Hardware configuration example of the input support device 100) Next, with reference to FIG. 2, a hardware configuration example of the input support device 100 will be described. Here, the case where the input support device 100 is applied to a computer such as a PC or a tablet PC will be taken as an example for explanation. However, the input support device 100 may also be applied to a server that can be connected from a PC or the like used by the user.

[0046] FIG. 2 is a block diagram showing a hardware configuration example of the input support device 100 according to the embodiment. In FIG. 2, the input support device 100 includes a CPU (Central Processing Unit) 201, a ROM (Read Only Memory) 202, and a RAM (Random Access Memory) 203.

[0047] In addition, the input support device 100 includes a HDD (Hard Disc Drive) 204, an HD 205, a CD (Compact Disc)-RW (ReWritable) drive 206, and a CD-RW 207. The input support device 100 also includes a display 208, a keyboard 209, a mouse 210, and a network I / F (Interface) 211. Each component is connected by a bus 200 respectively.

[0048] Here, the CPU 201 controls the overall operation of the input support device 100. The CPU 201 may have multiple cores. The ROM 202 and the HD 205 store various programs. The programs stored in the ROM 202 and the HD 205 are, for example, this input support program. This input support program is applied to, for example, kana-kanji conversion software. The program stored in the ROM 202 is loaded into the CPU 201, causing the CPU 201 to execute the encoded processing. The RAM 203 is used as the work area of the CPU 201.

[0049] The HDD 204 controls the reading or writing of data to / from the HD 205 according to the control of the CPU 201. The HD 205 stores the data written according to the control of the HDD 204. The CD-RW drive 206 controls the reading or writing of data to / from the CD-RW 207 according to the control of the CPU 201. The CD-RW 207 stores the data written according to the control of the CD-RW drive 206. The CD-RW 207 may be detachable from the input support device 100, for example.

[0050] The display 208 displays various data such as a cursor, icons, menus, windows, toolboxes, characters, images, or function information. The display 208 is, for example, a liquid crystal display, an organic EL (Electroluminescence) display, etc.

[0051] The keyboard 209 has keys for inputting characters, numerical values, various instructions, etc., and performs data input. The mouse 210 selects or executes various instructions, selects a processing target, or moves the mouse pointer, etc. Further, the display 208 may be a touch panel and may have functions corresponding to the keyboard 209 and functions corresponding to the mouse 210. In this case, the input support device 100 may not have the keyboard 209 and the mouse 210.

[0052] The network I / F 211 is connected to the network NW through a communication line and is connected to other computers via the network NW. The network NW is, for example, a LAN (Local Area Network), a WAN (Wide Area Network), the Internet, etc. The network I / F 211 manages the interface between the network NW and the inside of the input support device 100, and controls the input and output of data from other computers. The network I / F 211 is, for example, a modem, a LAN adapter, etc.

[0053] In addition to the components described above, the input support device 100 may have, for example, a DVD (Digital Versatile Disc) drive, an SSD (Solid State Drive), a USB (Universal Serial Bus) port, etc. Further, in addition to the components described above, the input support device 100 may have, for example, a printer, a scanner, a microphone, or a speaker, etc. Further, the input support device100 may not have, for example, the HDD 204, the HD 205, the CD-RW drive 206, the CD-RW 207, etc. among the components described above.

[0054] (Stored content of the semantic connection table 300) Next, with reference to FIG. 3, the stored content of the semantic connection table 300 used by the input support device 100 will be described. The semantic connection table 300 is realized by a storage device such as the RAM 203, the HD 205, etc.

[0055] FIG. 3 is an explanatory diagram showing an example of the stored content of the semantic connection table 300. In FIG. 3, the semantic connection table 300 has fields for a previous word, a subsequent word, and a connection strength, and stores semantic connection information (for example, semantic connection information 300-1 to 300-3) as records by setting information in each field.

[0056] Here, the previous word is a word (kana) that has a semantic connection with the subsequent word that appears after the previous word. The subsequent word is a word (kana) that has a semantic connection with the previous word that appears before the subsequent word. The previous word and the subsequent word are, for example, continuously input during Japanese input, or are input with a particle or auxiliary verb such as "te ni o wa" in between. The previous word and the subsequent word may be associated with a conversion result obtained by converting the word (kana) into Chinese characters.

[0057] The connection strength indicates the strength (degree) of the semantic connection between the previous word and the subsequent word. The connection strength is represented by a numerical value from 0 to 1, for example, and the larger the value, the stronger the semantic connection. However, the connection strength may be represented by levels such as "strong, medium, weak".

[0058] For example, the semantic connection information 300-1 indicates the semantic connection strength "0.78" between the previous word "ikken" and the subsequent word "rakuchaku" that are continuously input.

[0059] In the example of FIG. 3, the front-back relationship between two words with a semantic connection is used as an example for explanation, but it is not limited to this. For example, the semantic connection table 300 may store the front-back relationship (word order) between three or more words with a semantic connection. Specifically, for example, the semantic connection table 300 may store the front-back relationship between three words (previous word, middle word, subsequent word) with a semantic connection, and the strength of the semantic connection between the three words.

[0060] The semantic connection table 300 may be automatically created by analyzing Japanese document data, for example, or may be created manually. Also, the stored content of the semantic connection table 300 may be updated at any time. For example, according to the number of occurrences or the occurrence frequency of combinations of words having a semantic connection in Japanese document data, a new semantic connection between words may be registered or the connection strength may be updated. The connection strength is updated so that, for example, the higher the occurrence frequency of a combination of words having a semantic connection, the stronger it becomes.

[0061] (Functional configuration example of the input support device 100) Next, a functional configuration example of the input support device 100 will be described with reference to FIG. 4.

[0062] FIG. 4 is a block diagram showing a functional configuration example of the input support device 100 according to the embodiment. In FIG. 4, the input support device 100 includes a reception unit 401, a specification unit 402, a candidate creation unit 403, an extraction unit 404, a determination unit 405, an input error detection unit 406, an output unit 407, and a storage unit 410. The reception unit 401 to the output unit 407 have functions as a control unit. For example, by causing the CPU 201 to execute a program stored in a storage device such as the ROM 202, the RAM 203, or the HD 205 shown in FIG. 2, or by means of the network I / F 211, their functions are realized. The processing results of each functional unit are stored in a storage device such as the RAM 203 or the HD 205, for example. The storage unit 410 is realized by a storage device such as the RAM 203 or the HD 205, for example. Specifically, for example, the storage unit 410 stores the semantic connection table 300 shown in FIG. 3. The storage unit 110 shown in FIG. 1 corresponds to the storage unit 410, for example.

[0063] The reception unit 401 receives an input of a character string. The character string to be input is, for example, a character string to be subject to kana-kanji conversion. Specifically, for example, the reception unit 401 receives an input of a character string by a user operation using the keyboard 209 and the mouse 210 shown in FIG. 2. For example, in the case of Japanese input in the Roman character input mode, the character string to be input corresponds to Roman character spelling or the result of converting Roman character spelling into kana. Also, in the case of Japanese input in the kana input mode, the character string to be input corresponds to kana spelling (keystrokes).

[0064] Based on the character sequence of the input character string, the specifying unit 402 specifies, from the input character string, a portion where the value indicating the unnaturalness of the character sequence is smaller than the first reference value and equal to or greater than the second reference value. The input character string is a character string before confirmation. The value indicating the unnaturalness of the character sequence is an index value for evaluating the character sequence, and the higher the value, the more unnatural the character sequence is.

[0065] The first reference value is a value for detecting an input error, and is set to a value that can be said to deviate from the natural character sequence, for example. The second reference value is a value smaller than the first reference value, and is set to a value that can be said that the character sequence is somewhat unnatural but not detected as an input error, for example.

[0066] Input errors include, for example, insertion errors, deletion errors, substitution errors, and transposition errors. An insertion error is an error in which an arbitrary keystroke character is input additionally (e.g., nyuuryoku (にゅうりょく) → nyuuuryoku (にゅううりょく)). A deletion error is an error in which an arbitrary keystroke character is missing (e.g., nyuuryoku (にゅうりょく) → nyuuyoku (にゅうよく)).

[0067] A substitution error is an error in which an arbitrary keystroke character is mistaken for another character (such as a character on a neighboring key) (e.g., nyuuryoku (にゅうりょく) → nuuuryoku (ぬううりょく)). A transposition error is an error in which the keystroke order of an arbitrary keystroke character is reversed (e.g., nyuuryoku (にゅうりょく) → nyuuyroku (にゅうyろく)).

[0068] In the following description, the value indicating the unnaturalness of the character sequence may be referred to as the "unnaturalness score". Also, in the input string, a portion where the unnaturalness score is less than the first reference value and equal to or greater than the second reference value may be referred to as a "suspected portion". The suspected portion corresponds to a portion where an input error is suspected.

[0069] Specifically, for example, the specific part 402 divides the input string into elements. The elements are, for example, clauses or words. However, it is arbitrarily possible to set by what unit the input string is divided. The elements may include portions where input errors have occurred. Next, the specific part 402 calculates an unnaturalness score for each divided element based on the character sequence of the input string. Then, the specific part 402 specifies, as a suspected portion, an element for which the calculated unnaturalness score is less than the first reference value and equal to or greater than the second reference value.

[0070] The unnaturalness score may be calculated using any existing technique. For example, the unnaturalness score is calculated based on the plausibility of the transition between adjacent characters. The adjacent characters are, for example, adjacent Roman characters in the case of Japanese input in Roman character input mode. Also, in the case of Japanese input in kana input mode, the adjacent characters are adjacent kana characters.

[0071] The plausibility of the transition between adjacent characters is represented, for example, by the transition probability between adjacent characters. The transition probability between adjacent characters is derived, for example, based on the sequence of correct characters (Roman characters or kana characters) input in the past during Japanese input. Also, the unnaturalness score of an element may be calculated based on the plausibility of the first character of that element being at the beginning. Also, the unnaturalness score of an element may be calculated based on the plausibility of the last character of that element being at the end.

[0072] Also, the unnaturalness score of an element may be calculated based on the transition probability from the character immediately preceding the element to the first character of the element in the input string. Also, the unnaturalness score of an element may be calculated based on the transition probability from the last character of the element to the character immediately following the element in the input string.

[0073] The candidate creation unit 403 creates reading candidates for the specified suspicious part by performing a predetermined correction process on the specified suspicious part. Specifically, for example, the candidate creation unit 403 performs a correction process on the suspicious part, such as supplementing characters, deleting unnecessary characters, or swapping the order of characters, so that the learned Roman characters (or kana characters) in Japanese are in a natural order. Thereby, the candidate creation unit 403 creates one or more reading candidates for the suspicious part.

[0074] The extraction unit 404 extracts at least one of the words before and after the specified suspicious part from the input string. Specifically, for example, the extraction unit 404 extracts the word immediately before the specified suspicious part from the input string. Also, the extraction unit 404 extracts the word immediately after the specified suspicious part from the input string.

[0075] However, particles and auxiliary verbs such as "てにをは" may be excluded from the extraction targets. For example, when the word immediately before the suspicious part is a particle, the extraction unit 404 extracts the word immediately before that particle. Also, when the word immediately after the suspicious part is a particle, the extraction unit 404 extracts the word immediately after that particle. Also, the extraction unit 404 may extract two or more words before the suspicious part. Also, the extraction unit 404 may extract two or more words after the suspicious part. In this case, the maximum number of words to be extracted may be set in advance.

[0076] The determination unit 405 refers to the storage unit 410 and determines whether there is a reading candidate among the created reading candidates that has a semantic connection with at least one of the words before and after the suspicious part in the input string. Here, the sequential relationship between words with a semantic connection that are continuously input is stored.

[0077] For example, assume that the word before the suspected part is extracted from the input string. In this case, the determination unit 405 refers to, for example, the semantic connection table 300 shown in FIG. 3, and checks whether there is a reading candidate (corresponding to the "subsequent word") that appears after the extracted word (corresponding to the "previous word") and has a semantic connection with the extracted word among the created reading candidates.

[0078] Also, assume that the word after the suspected part is extracted from the input string. In this case, the determination unit 405 refers to, for example, the semantic connection table 300, and checks whether there is a reading candidate (corresponding to the "previous word") that appears before the extracted word (corresponding to the "subsequent word") and has a semantic connection with the extracted word among the created reading candidates.

[0079] Furthermore, the storage unit 410 may further store the strength of the semantic connection between continuously input words. In this case, when the determination unit 405 determines that there is a reading candidate with a semantic connection, it may refer to the storage unit 410 to specify the strength of the semantic connection between the reading candidate and at least one of the words before and after the suspected part.

[0080] The strength of the semantic connection is represented by, for example, a numerical value from 0 to 1, and the larger the value, the stronger the semantic connection. Also, the strength of the semantic connection may be represented by levels such as "strong, medium, weak", for example.

[0081] As an example, assume that the word before the suspected part is extracted from the input string, and there is a reading candidate that appears before the extracted word and has a semantic connection with the extracted word among the created reading candidates. In this case, the determination unit 405 refers to, for example, the semantic connection table 300 to specify the strength of the semantic connection between the reading candidate having a semantic connection with the extracted word and the extracted word.

[0082] The input error detection unit 406 detects input errors from the input string. When it is determined that there is a reading candidate that has a semantic connection with at least one of the words before and after the suspected location, the input error detection unit 406 detects an input error at the suspected location. On the other hand, when it is determined that there is no reading candidate that has a semantic connection with at least one of the words before and after the suspected location, the input error detection unit 406 does not detect an input error at the suspected location.

[0083] Also, when it is determined that there is a reading candidate with a semantic connection and the strength of the identified semantic connection is equal to or greater than the threshold α, the input error detection unit 406 may detect an input error at the suspected location. On the other hand, even when it is determined that there is a reading candidate with a semantic connection, if the strength of the identified semantic connection is less than the threshold α, the input error detection unit 406 may decide not to detect an input error at the suspected location.

[0084] The threshold α can be arbitrarily set. For example, when the strength of the semantic connection is represented by a numerical value between 0 and 1, the threshold α may be set to a value of about 0.5. Also, when the strength of the semantic connection is represented by the levels of "strong, medium, weak", the threshold α may be set to "medium".

[0085] When an input error at the suspected location is detected, the output unit 407 outputs a conversion candidate corresponding to the input string based on the created reading candidates. Here, the conversion candidate is, for example, the reading corresponding to the input string converted into Chinese characters or a sentence containing Chinese characters.

[0086] Specifically, for example, first, the output unit 407 creates a reading corresponding to the input string based on the created reading candidates. The reading corresponding to the input string is created, for example, by using the reading of the suspected location in the input string as the created reading candidate. Next, the output unit 407 refers to the mapping information associating readings with Chinese characters and creates a conversion candidate corresponding to the input string by performing kana-to-Chinese-character conversion on the created reading.

[0087] Then, the output unit 407 outputs the created conversion candidates. To explain in more detail, for example, the output unit 407 displays the created conversion candidates on the display 208 so that they can be selected for the input character string. At this time, the output unit 407 may display the created conversion candidates together with a message indicating that the input error has been corrected.

[0088] In addition, when there are a plurality of created reading candidates, conversion candidates corresponding to each of the plurality of reading candidates are created. In this case, the output unit 407 may preferentially output the conversion candidates corresponding to the reading candidates that have a semantic connection with at least one of the words before and after the suspected part among the created conversion candidates.

[0089] Also, when an input error is detected at the suspected part, the output unit 407 may output conversion candidates corresponding to the input character string based on the created reading candidates and the original reading of the suspected part. The original reading of the suspected part is the reading corresponding to the arrangement of the characters in the suspected part.

[0090] Specifically, for example, the output unit 407 creates a first conversion candidate corresponding to the input character string based on the created reading candidates. Note that when there are a plurality of created reading candidates, there are a plurality of first conversion candidates. Also, the output unit 407 creates a second conversion candidate corresponding to the input character string based on the original reading of the suspected part.

[0091] Then, the output unit 407 outputs the created first conversion candidate and second conversion candidate. At this time, the output unit 407 may preferentially output the conversion candidates (corresponding to any of the first conversion candidates) corresponding to the reading candidates that have a semantic connection with at least one of the words before and after the suspected part among the created first conversion candidate and second conversion candidate.

[0092] Further, the output unit 407 may set a priority for each of the created conversion candidates (for example, the first conversion candidate and the second conversion candidate). Here, the priority indicates the degree of priority as the correct reading of the suspected part. Then, the output unit 407 may output the created conversion candidates according to the set priorities. More specifically, for example, the output unit 407 may display the created conversion candidates on the display 208 in descending order of the set priorities.

[0093] Here, the specific processing content when setting priorities for each of the created first conversion candidate and second conversion candidate will be described.

[0094] First, the output unit 407 sets a priority for the created first conversion candidate according to the unnaturalness score of the reading candidate corresponding to the first conversion candidate. The priority is set, for example, so that the lower the unnaturalness score, the higher the priority. However, the priority for each conversion candidate (for example, the first conversion candidate) is set in consideration of the priority for other conversion candidates (for example, the second conversion candidate).

[0095] The unnaturalness score of the reading candidate is a value indicating the unnaturalness of the arrangement of the characters (for example, Romanization) corresponding to the reading candidate. The unnaturalness score of the reading candidate is calculated based on, for example, the arrangement of the characters corresponding to the reading candidate, the characters before and after the suspected part in the input string, and the like.

[0096] In addition, the output unit 407 sets a priority for the created second conversion candidate according to the unnaturalness score of the original reading. The unnaturalness score of the original reading is a value indicating the unnaturalness of the arrangement of the characters at the suspected part (the part corresponding to the original reading). The unnaturalness score of the original reading is calculated based on, for example, the arrangement of the characters at the suspected part, the characters before and after the suspected part in the input string, and the like.

[0097] Further, among the created first conversion candidates and second conversion candidates, the output unit 407 may set the highest priority for a conversion candidate (corresponding to any of the first conversion candidates) that has a semantic connection with at least one of the words before and after the suspected part.

[0098] Then, the output unit 407 outputs the created first conversion candidates and second conversion candidates according to the set priorities. More specifically, for example, the output unit 407 may display the first conversion candidates and second conversion candidates on the display 208 in descending order of the set priorities.

[0099] An output example of the conversion candidates corresponding to the input character string will be described later with reference to FIGS. 6 and 7.

[0100] Also, the input error detection unit 406 may detect input errors at positions (for example, elements) in the input character string where the unnaturalness score is equal to or higher than the first reference value. In this case, the output unit 407 may create a reading candidate for the position corresponding to the input error in the input character string, for example, by performing a predetermined correction process on the input error. Then, the output unit 407 may output a conversion candidate corresponding to the input character string based on the created reading candidate for the position corresponding to the input error.

[0101] Note that the functional units (reception unit 401 to output unit 407) of the input support device 100 may be realized by, for example, a server that can be connected via a network NW (see FIG. 2) from a PC or the like used by the user. In this case, the input support device 100 receives an input of a character string from the PC used by the user by the reception unit 401, and outputs the created conversion candidates to the PC used by the user by the output unit 407.

[0102] Further, the functional units (reception unit 401 to output unit 407) of the input support device 100 may be realized by a plurality of computers (for example, a server and a PC). For example, the specifying unit 402 to the input error detection unit 406 may be realized by a server, and the reception unit 401 and the output unit 407 may be realized by a PC. In this case, the communication between the functional units of different computers is performed, for example, by transmission and reception between the functional units via the network NW.

[0103] (Example of creating conversion candidates corresponding to the input string) Next, with reference to FIG. 5, an example of creating conversion candidates corresponding to the input string when an input error is detected at the suspected location will be described.

[0104] FIG. 5 is an explanatory diagram showing an example of creating conversion candidates corresponding to the input string. In FIG. 5, it is assumed that a string 501 "ichouisseki" is input by a user's operation input using the keyboard 209 and the mouse 210 shown in FIG. 2. "ichouisseki" is the input Romanized spelling. "いちょういっせき" is the result of converting "ichouisseki" into kana.

[0105] The specifying unit 402 specifies a suspected location where the unnaturalness score is less than the first reference value and equal to or greater than the second reference value from the input string 501. Here, it is assumed that the input string 501 is divided into "ichou" and "isseki".

[0106] Also, assume that the unnaturalness score of "ichou" is less than the first reference value and equal to or greater than the second reference value. Also, assume that the unnaturalness score of "isseki" is less than the second reference value. In this case, the specifying unit 402 specifies a suspected part 502 from the input character string 501. The suspected part 502 corresponds to "ichou". Although the unnaturalness score of the suspected part 502 is not so high as to deviate from the natural order of characters, it is a part where the unnaturalness score is high enough to suspect an input error.

[0107] The candidate creation unit 403 creates a reading candidate 504 for the specified suspected part 502 by performing a predetermined correction process on the specified suspected part 502. Here, assume that the reading candidate 504 is created by supplementing "c" after "i" in the suspected part 502 and converting it into kana so as to be in the natural order of the learned Roman characters in Japanese.

[0108] The extraction unit 404 extracts at least one of the words before and after the specified suspected part 502 from the input character string 501. Here, the word 503 "isseki" immediately after the suspected part 502 is extracted from the input character string 501.

[0109] The determination unit 405 refers to the semantic connection table 300 and determines whether there is a reading candidate among the created reading candidates 504 that has a semantic connection with the extracted word 503. Here, the word 503 matches the subsequent word "isseki" of the semantic connection information 300-2 in the semantic connection table 300.

[0110] Also, the reading candidate 504 matches the previous word "itchou" of the semantic connection information 300-2. Therefore, it can be said that the reading candidate 504 is a reading candidate that has a semantic connection with the extracted word 503. For this reason, the determination unit 405 determines that there is a reading candidate among the created reading candidates 504 that has a semantic connection with the extracted word 503.

[0111] Next, the input error detection unit 406 determines whether to detect an input error at the suspected location 502. Here, it is assumed that an input error at the suspected location 502 is detected when it is determined that there is a reading candidate having a semantic connection with the extracted word 503 and the strength of the semantic connection is equal to or greater than the threshold α.

[0112] Here, the threshold α is set to "α = 0.5". Also, the strength of the semantic connection between the created reading candidate 504 and the extracted word 503 is "0.67" from the semantic connection information 300-2. In this case, since there is a reading candidate 504 having a semantic connection with the extracted word 503 and the strength of the semantic connection is equal to or greater than the threshold α, the input error detection unit 406 detects an input error at the suspected location 502.

[0113] When an input error at the suspected location 502 is detected, the output unit 407 outputs a conversion candidate corresponding to the input character string 501 based on the created reading candidate 504. Specifically, for example, the output unit 407 creates a reading 505 corresponding to the input character string 501 by setting the reading of the suspected location 502 in the input character string 501 to the created reading candidate 504. Here, the reading 505 is "itchou iseki".

[0114] Then, the output unit 407 creates a conversion candidate 506 corresponding to the input character string 501 by performing kana-kanji conversion on the created reading 505. The kana-kanji conversion is performed using mapping information (not shown) that associates readings with kanji. Here, "ichou ichiya" is created as the conversion candidate 506.

[0115] Thereby, the input support device 100 can detect an input error at the suspected location 502 by using the semantic connection with the surrounding words (word 503). Also, the input support device 100 can repair the suspected location 502 so as to have a natural arrangement of Roman characters as Japanese and output the conversion candidate 506.

[0116] For example, suppose that the suspected portion 502 is a portion where the user mistakenly inputs "ichou" when trying to input the word "ichiha ichiyatsu." The input support device 100 can repair the suspected portion 502 and output a conversion candidate 506 desired by the user.

[0117] (Example of conversion candidate output) Next, an example of output of conversion candidates corresponding to an input character string will be described with reference to Fig. 6 and Fig. 7. Here, an example of output of conversion candidates will be described by taking the input character string 501 shown in Fig. 5 as an example.

[0118] 6 is an explanatory diagram (part 1) showing an output example of conversion candidates. In (6-1) of FIG. 6, a character string 501 is input into a document 600 by a user operation. The user operation is performed, for example, using the keyboard 209 or the mouse 210 shown in FIG. 2. The document 600 is displayed, for example, on the display 208 shown in FIG. 2.

[0119] In the input character string 501, an input error is detected in a suspected portion 502 (see FIG. 5). Here, a case will be described in which a conversion candidate corresponding to the input character string 501 is generated based on the generated reading candidate 504 (see FIG. 5).

[0120] Specifically, a reading "icchou isseki" corresponding to the input character string 501 is generated based on the generated reading candidate 504 (see FIG. 5). The kana-kanji conversion process is executed, for example, in response to a conversion instruction from the user. The conversion instruction is issued, for example, by pressing a specific key on the keyboard 209.

[0121] In (6-2) of Fig. 6, conversion candidates 601 corresponding to the input character string 501 are displayed in a selectable manner based on the created reading "icchou isseki." The conversion candidate 601 is "ichiasa ichiyaku." Among the conversion candidates 601, a portion 602 (completed) where correction processing has been performed for an input mistake is displayed in a different manner from the other portions.

[0122] Here, the background color of part 602 is different from that of other parts. As another aspect, for example, part 602 may be made bold or underlined. Also, a message 603 is displayed in association with the conversion candidate 601. The message 603 is a message indicating that an input error included in the input string 501 has been corrected.

[0123] As a result, even when the user makes an input error during Japanese input, the user can obtain the desired conversion candidate 601 without re - entering. Also, by referring to the message 603, the user can grasp that the input error has been automatically corrected. Further, the user can grasp that the part 602 with a background color different from other parts among the conversion candidates 601 is the corrected part.

[0124] Note that after the conversion candidate 601 is displayed, for example, when a confirmation instruction (selection operation) is performed by the user's operation, the conversion candidate 601 is selected and the input is confirmed.

[0125] FIG. 7 is an explanatory diagram (part 2) showing an output example of conversion candidates. In (7 - 1) of FIG. 7, a string 501 is input into the document 700 by the user's operation. In the input string 501, an input error at the suspected part 502 (see FIG. 5) is detected. Here, a case will be described where a conversion candidate corresponding to the input string 501 is created based on the created reading candidate 504 (see FIG. 5) and the original reading of the suspected part 502.

[0126] Specifically, based on the created reading candidate 504 (see FIG. 5), a reading "itchou isseki" corresponding to the input string 501 is created. Also, based on the original reading of the suspected part 502, a reading "ichou isseki" corresponding to the input string 501 is created.

[0127] In (7-2) of Fig. 7, based on the created reading "iccho isseki" and the reading "iccho isseki", a first conversion candidate 701 and a second conversion candidate 702 corresponding to the input character string 501 are displayed in a selectable manner. The first conversion candidate 701 is "ichiasa ichiyaku". The second conversion candidate 702 is "icho isseki".

[0128] Here, the highest priority is set to the first conversion candidate 701 corresponding to the reading candidate 504 among the first conversion candidate 701 and the second conversion candidate 702. Therefore, the first conversion candidate 701 is displayed higher than the second conversion candidate 702. Also, a message 703 is displayed in association with the first conversion candidate 701. The message 703 is a message indicating that an input error included in the input character string 501 has been corrected.

[0129] The user can obtain a first conversion candidate 701 and a second conversion candidate 702 for a portion where an input error is suspected. As a result, if the user makes an input error when inputting Japanese, the user can obtain the first conversion candidate 701 without having to re-input the text. Also, by referring to message 703, the user can understand that the input error has been automatically corrected. Also, if the user made no input error when inputting Japanese (intentionally inputting "ichou isseki"), the user can obtain the second conversion candidate 702 that corresponds to the original reading of the suspected portion 502.

[0130] In addition, after the first conversion candidate 701 and the second conversion candidate 702 are displayed, when, for example, a user operates to select the first conversion candidate 701 or the second conversion candidate 702, the first conversion candidate 701 or the second conversion candidate 702 is selected and the input is confirmed.

[0131] (Input Support Processing Procedure of the Input Support Device 100) Next, the input support process procedure of the input support device 100 will be described with reference to FIGS.

[0132] FIG. 8 and FIG. 9 are flowcharts showing an example of the input support processing procedure of the input support device 100 according to the embodiment. In the flowchart of FIG. 8, first, the input support device 100 determines whether an input error is detected from the input string (step S801). In step S801, the input support device 100 detects an input error at a location (element) in the input string where the unnaturalness score is equal to or greater than the first reference value.

[0133] Here, when an input error is detected (step S801: Yes), the input support device 100 creates a reading candidate for the location corresponding to the input error in the input string by performing a predetermined correction process on the input error (step S802).

[0134] Next, the input support device 100 creates a conversion candidate corresponding to the input string based on the reading candidate for the location corresponding to the created input error (step S803). Specifically, for example, the input support device 100 creates a reading corresponding to the input string by using the reading of the location corresponding to the input error in the input string as the created reading candidate, and creates a conversion candidate corresponding to the input string by performing kana-kanji conversion on the created reading.

[0135] Then, the input support device 100 outputs the created conversion candidate for the input string (step S804), and ends the series of processes according to this flowchart.

[0136] Also, in step S801, when no input error is detected (step S801: No), the input support device 100 identifies a suspected location in the input string where the unnaturalness score is less than the first reference value and equal to or greater than the second reference value based on the character order of the input string (step S805).

[0137] Then, the input support device 100 determines whether a suspicious part has been identified (step S806). Here, if no suspicious part has been identified (step S806: No), the input support device 100 creates conversion candidates corresponding to the input character string based on the original reading of the input character string (step S807). Then, the input support device 100 outputs the created conversion candidates for the input character string (step S808), and ends the series of processes according to this flowchart.

[0138] Also, in step S806, if a suspicious part has been identified (step S806: Yes), the input support device 100 proceeds to step S901 shown in FIG. 9.

[0139] In the flowchart of FIG. 9, first, the input support device 100 creates reading candidates for the identified suspicious part by performing a predetermined correction process on the identified suspicious part (step S901). Next, the input support device 100 extracts at least one of the words before and / or after the identified suspicious part from the input character string (step S902).

[0140] Then, the input support device 100 determines whether the word before the suspicious part has been extracted (step S903). Here, if the previous word has not been extracted (step S903: No), the input support device 100 proceeds to step S905.

[0141] On the other hand, if the previous word has been extracted (step S903: Yes), the input support device 100 refers to the semantic connection table 300 and determines whether there is a reading candidate (corresponding to the "subsequent word") that appears after the extracted word (corresponding to the "previous word") and has a semantic connection with the extracted word among the reading candidates created in step S901 (step S904).

[0142] Here, if there is a reading candidate with a semantic connection (step S904: Yes), the input support device 100 proceeds to step S907. On the other hand, if there is no reading candidate with a semantic connection (step S904: No), the input support device 100 determines whether a word after the suspicious part has been extracted (step S905).

[0143] Here, if the subsequent word has not been extracted (step S905: No), the input support device 100 returns to step S807 shown in FIG. 8. On the other hand, if the subsequent word has been extracted (step S905: Yes), the input support device 100 refers to the semantic connection table 300 and determines whether there is a reading candidate (corresponding to the "previous word") that appears before the extracted word (corresponding to the "subsequent word") and has a semantic connection with the extracted word among the reading candidates created in step S901 (step S906).

[0144] Here, if there is no reading candidate with a semantic connection (step S906: No), the input support device 100 returns to step S807 shown in FIG. 8. On the other hand, if there is a reading candidate with a semantic connection (step S906: Yes), the input support device 100 detects an input error at the specified suspicious part (step S907).

[0145] Next, the input support device 100 creates a conversion candidate corresponding to the input character string based on the created reading candidates for the suspicious part (step S908). Specifically, for example, the input support device 100 creates a reading corresponding to the input character string by using the reading of the suspicious part as the created reading candidate among the input character string, and creates a conversion candidate corresponding to the input character string by performing kana-kanji conversion on the created reading.

[0146] Then, the input support device 100 outputs the created conversion candidate for the input character string (step S909) and ends the series of processes according to this flowchart.

[0147] As a result, the input support device 100 can output conversion candidates corresponding to the input character string. For example, when an input error is detected from the input character string, the input support device 100 can output conversion candidates corresponding to the input character string based on the reading candidates at that location.

[0148] In addition, in step S902, when the words before and after the suspected location are extracted, the input support device 100 may determine whether there is a reading candidate that appears after the extracted previous word and before the extracted next word among the reading candidates created in step S901 and has a semantic connection with the extracted previous word and the next word. Then, when there is a reading candidate with a semantic connection, the input support device 100 may detect an input error at the specified suspected location. On the other hand, when there is no reading candidate with a semantic connection, the input support device 100 may not detect an input error at the specified suspected location. In this case, the input support device 100 returns to step S807 shown in FIG. 8.

[0149] As described above, according to the input support device 100 according to the embodiment, based on the arrangement of the characters in the input character string, a suspicious location where the unnaturalness score is smaller than the first reference value for detecting an input error and not less than a second reference value smaller than the first reference value can be specified. Further, according to the input support device 100, by performing a predetermined correction process on the specified suspicious location, reading candidates for the suspicious location are created, and referring to the storage unit 410, it can be determined whether there is a reading candidate among the created reading candidates that has a semantic connection with at least one of the words before and after the suspicious location in the input character string. The storage unit 410 stores the context between continuously input words with a semantic connection. Then, according to the input support device 100, when there is a reading candidate that has a semantic connection with at least one of the words before and after, an input error at the suspicious location can be detected.

[0150] As a result, when detecting input errors from the input string, the input assistance device 100 can improve the detection accuracy of input errors by the user for the locations (suspected locations) where input errors are suspected, by utilizing the semantic connection with the surrounding words. For example, even when the unnaturalness score is not high enough to say that it deviates from the natural arrangement of characters, if there is a semantic connection between the reading candidates and the surrounding words, the input assistance device 100 can determine that it is not the arrangement of characters that was originally intended to be input, and detect input errors at the suspected locations.

[0151] Moreover, according to the input assistance device 100, when detecting an input error at a suspected location, it can output conversion candidates corresponding to the input string based on the created reading candidates for the suspected location.

[0152] As a result, the input assistance device 100 can enhance the accuracy of outputting the conversion candidates desired by the user by outputting conversion candidates based on the reading candidates with the input errors repaired for the locations (suspected locations) where input errors are detected.

[0153] Furthermore, according to the input assistance device 100, it can output conversion candidates corresponding to the input string based on the created reading candidates for the suspected location and the original reading of the suspected location.

[0154] As a result, the input assistance device 100 can further enhance the accuracy of outputting the conversion candidates desired by the user for the locations (suspected locations) where input errors are detected, by considering the case where there are no input errors, and outputting conversion candidates not only based on the reading candidates with the input errors repaired but also based on the original reading.

[0155] In addition, according to the input support device 100, based on the created reading candidates of the suspected part, a first conversion candidate corresponding to the input character string can be created, and based on the original reading of the suspected part, a second conversion candidate corresponding to the input character string can be created. Then, according to the input support device 100, among the created first conversion candidate and second conversion candidate, the conversion candidate (corresponding to any of the first conversion candidates) corresponding to the reading candidate having a semantic connection with at least one of the words before and after the suspected part can be preferentially output.

[0156] Thereby, when the input support device 100 outputs a plurality of conversion candidates corresponding to the input character string, by preferentially outputting the ones that are likely to be the conversion candidates desired by the user, the convenience when performing a confirmation instruction (selection operation) by the user can be improved.

[0157] In addition, according to the input support device 100, by referring to the storage unit 410 in which the strength of the semantic connection between words is stored, when there is a reading candidate having a semantic connection with at least one of the words before and after the suspected part, and the strength of the semantic connection is equal to or greater than the threshold α, an input error at the suspected part can be detected.

[0158] Thereby, for the part where an input error is suspected (the suspected part), even if there is a semantic connection with the surrounding words, when the semantic connection is weak, the input error at the suspected part can be prevented from being detected. This detection method is useful, for example, when the stored content of the semantic connection table 300 (for example, the connection strength between words) is updated at any time.

[0159] From these, according to the input support program, input support method, and input support device according to the embodiment, input errors by the user can be accurately detected, and the conversion accuracy of Japanese input can be improved. For example, when creating a document using a document creation application, the trouble of the user re - inputting can be reduced, the work load related to Japanese input can be reduced, and the efficiency of document creation can be improved.

[0160] Note that the input support method described in this embodiment can be realized by executing a pre-prepared program on a computer such as a personal computer or a workstation. This input support program is recorded on a computer-readable recording medium such as a hard disk, a flexible disk, a CD-ROM, a DVD, or a USB memory, and is executed by being read from the recording medium by the computer. Further, this input support program may be distributed via a network such as the Internet.

Industrial Applicability

[0161] The input support program, input support method, and input support device according to this invention are useful for a computer system that supports character input based on a user's operation, and in particular, are suitable for a computer system that inputs Japanese based on a user's operation.

Explanation of Signs

[0162] 100 Input support device 101, 501 Character string 102 Location 103, 503 Word 104, 504 Reading candidates 110, 410 Storage unit 200 Bus 201 CPU 202 ROM 203 RAM 204 HDD 205 HD 206 CD-RW drive 207 CD-RW 208 Display 209 Keyboard 210 Mouse 211 Network I / F 300 Table 401 Reception unit 402 Identification unit 403 Candidate creation unit 404 Extraction Unit 405 Judgment Unit 406 Input Error Detection Unit 407 Output Unit 502 Suspicious Location 505 Reading 506, 601, 701, 702 Conversion Candidates 600, 700 Documents 602 Part 603, 703 Messages

Claims

1. Based on the character sequence of the input string, identify a location where the value indicating the unnaturalness of the character sequence in the string is less than a first reference value for detecting input errors and is equal to or greater than a second reference value smaller than the first reference value. Create a reading candidate for the identified location by performing a predetermined correction process on the identified location. Refer to a storage unit that stores the context between continuously input words with semantic connections, and determine whether there is a reading candidate among the created reading candidates that has a semantic connection with at least one of the words before and after the location in the string. If there is a reading candidate with a semantic connection to any of the words, detect an input error at the location. An input assistance program characterized by causing a computer to execute the process.

2. When detecting an input error at the location, output a conversion candidate corresponding to the string based on the created reading candidate. The input assistance program according to claim 1, characterized by causing the computer to execute the process.

3. The process of outputting Outputs a conversion candidate corresponding to the string based on the created reading candidate and the original reading according to the character sequence of the location. The input assistance program according to claim 2, characterized by this.

4. Create a first conversion candidate corresponding to the string based on the created reading candidate. Create a second conversion candidate corresponding to the string based on the original reading. Cause the computer to execute the process. The process of outputting Among the created first conversion candidate and the second conversion candidate, preferentially output a conversion candidate corresponding to a reading candidate having a semantic connection with any of the words. The input assistance program according to claim 3, characterized by this.

5. The storage unit further stores the strength of the semantic connection between the words. The process of detecting When there is a reading candidate with a semantic connection and the strength of the semantic connection between the reading candidate with a semantic connection and any of the words is equal to or greater than a threshold value, detect an input error at the location. The input assistance program according to claim 1, characterized by this.

6. Based on the character sequence of the input string, identify a location where a value indicating the unnaturalness of the character sequence in the string is less than a first reference value for detecting input errors and is equal to or greater than a second reference value that is less than the first reference value. Create a reading candidate for the identified location by performing a predetermined correction process on the identified location. Refer to a storage unit that stores the context between words with a semantic connection that are continuously input, and determine whether there is a reading candidate among the created reading candidates that has a semantic connection with at least one of the words before and after the location in the string. If there is a reading candidate that has a semantic connection with any of the words, detect an input error at the location. An input support method characterized in that a computer executes the process.

7. Based on the character sequence of the input string, identify a location where a value indicating the unnaturalness of the character sequence in the string is less than a first reference value for detecting input errors and is equal to or greater than a second reference value that is less than the first reference value. Create a reading candidate for the identified location by performing a predetermined correction process on the identified location. Refer to a storage unit that stores the context between words with a semantic connection that are continuously input, and determine whether there is a reading candidate among the created reading candidates that has a semantic connection with at least one of the words before and after the location in the string. If there is a reading candidate that has a semantic connection with any of the words, detect an input error at the location. An input support device characterized by having a control unit that executes the process.

Images