WO2014087704A1 - 入力支援装置、入力支援方法、及び入力支援プログラム - Google Patents
入力支援装置、入力支援方法、及び入力支援プログラム Download PDFInfo
- Publication number
- WO2014087704A1 WO2014087704A1 PCT/JP2013/071712 JP2013071712W WO2014087704A1 WO 2014087704 A1 WO2014087704 A1 WO 2014087704A1 JP 2013071712 W JP2013071712 W JP 2013071712W WO 2014087704 A1 WO2014087704 A1 WO 2014087704A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- input
- phrase
- language
- word
- character string
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/3332—Query translation
- G06F16/3337—Translation of the query language, e.g. Chinese to English
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/12—Use of codes for handling textual entities
- G06F40/126—Character encoding
- G06F40/129—Handling non-Latin characters, e.g. kana-to-kanji conversion
Definitions
- One aspect of the present invention relates to an input support device, an input support method, and an input support program.
- An input support apparatus includes a reception unit that receives an input character string in Roman characters, a first language phrase expressed in alphabet, and a second language corresponding to the first language phrase
- a search unit that executes a second process for searching for a phrase in the first language corresponding to the Roman character notation including the input character string, and a result of the first process and a result of the second process are input.
- An output unit for outputting as a candidate.
- An input support method is an input support method executed by an input support device, the receiving step of receiving an input character string in Roman letters, a phrase of a first language expressed in alphabet, A phrase of the first language including the input character string from a storage unit for storing the Roman language notation of the phrase of the second language corresponding to the phrase of the first language in a manner in which their correspondence can be specified
- a search step for executing a first process for searching for a word and a second process for searching for a phrase of the first language corresponding to the Roman alphabet notation including the input character string from the storage unit; And an output step for outputting the result of the second process and the result of the second process as input candidates.
- An input support program includes a reception unit that receives an input character string in Roman characters, a first language phrase expressed in alphabet, and a second language corresponding to the first language phrase
- a search unit for executing a second process for searching for a phrase in the first language corresponding to the Roman character notation including the input character string, and a result of the first process and a result of the second process are input.
- An output unit that outputs as a candidate is executed by a computer.
- the process of searching for a phrase from the input character string itself without translation and the process of translation based on the input character string are executed, and the phrase obtained from these processes (the first phrase) Language phrases) are collectively output as input candidates.
- the phrase obtained from these processes (the first phrase) Language phrases) are collectively output as input candidates.
- the input character string includes a plurality of words
- the search unit executes the first process and the second process for each of the plurality of words
- the output unit includes the plurality of words. May be output as input candidates.
- the input of the selected word / phrase may be confirmed by replacing the common first language word / phrase selected by the user with the entire input character string. . With this replacement, the user can immediately start the next editing operation.
- the input character string includes a plurality of words
- the output unit outputs, as input candidates, words in the first language searched for in some of the plurality of words. May be. By searching for input candidates using only some of the words, it is possible to avoid a situation in which too many or too few words in the first language are extracted.
- some words may be the last word in the input character string.
- the phrase of the first language selected by the user is replaced with only a word used as a search condition for the selected phrase among a plurality of words, The input of the selected word may be confirmed. By replacing in this way, the user can know which word has been replaced.
- an input candidate when an input candidate includes a word in a first language to which information indicating an unconfirmed part is added, and an input of a word having an unconfirmed part is confirmed, the input candidate is confirmed.
- a focus may be set on the fixed part. By setting the focus position in this way, the user can immediately start the next editing operation without performing an operation such as moving the cursor.
- an input candidate when an input candidate includes a word in a first language to which information indicating an unconfirmed part is added, and an input of a word having an unconfirmed part is confirmed, the input candidate is confirmed.
- word candidates that can be inserted into the undetermined part are presented, so that the user can create a sentence or a sentence more easily.
- the additional processing unit is unconfirmed by referring to a predetermined storage unit in which correspondence between an unconfirmed part of a word and a word that can be inserted into the unconfirmed part is stored. You may acquire the phrase which can be inserted in a part.
- the additional processing unit refers to a predetermined storage unit in which a corpus of at least one of the first language and the second language is stored, so that the unconfirmed part Words that can be inserted into the URL may be acquired.
- further input candidates may be displayed in an order according to the number of appearances in the corpus.
- the additional processing unit may estimate a phrase that can be inserted into an undetermined part by analyzing a confirmed character string positioned before the input character string.
- a word or phrase that can be inserted into an uncertain part based on a sentence or sentence that has already been input, a user who is inputting the sentence or sentence may be presented in advance. it can.
- the phrase estimated by the analysis of the confirmed character string may be displayed in preference to the phrase not estimated by the analysis.
- the additional processing unit acquires further input candidates for each undetermined part. And obtaining combination information indicating a natural combination of a phrase that can be inserted into the first unconfirmed part and a phrase that can be inserted into the second unconfirmed part, and further inputting the combination information into each unconfirmed part
- a word that can be output together with the candidate and inserted into the other unconfirmed part according to a user operation in one unconfirmed part of the first and second unconfirmed parts may be presented based on the combination information . In this way, by presenting a natural combination of phrases, the user can easily create a natural sentence or sentence.
- the input candidates may be displayed separately in a word column and a phrase column.
- the user can easily find and select a desired word / phrase.
- the phrases may be further divided into a plurality of columns and displayed according to the word meaning.
- the user can easily find and select a desired word / phrase by presenting the phrase after dividing it by word meaning.
- the input candidates may be displayed in a plurality of columns according to the meaning.
- the user can easily find and select a desired word / phrase by presenting the input candidate word / phrase after being divided into word meanings.
- input candidates may be displayed so that each column scrolls independently.
- the user can look around the words in a specific column while continuing to look at the words in other columns.
- the user can input a phrase of a target language without specifying the language of the input character string.
- alphabet in this specification means a general term for Western European characters arranged in a certain order.
- Roman character means a character used when a certain language is written in Latin. In this specification, expressing a certain language using Roman letters is called "Roman alphabet notation”.
- This system 1 is a computer system that supports English text input. By using this system 1, a user who is not familiar with English can expect to easily create English sentences.
- the user may input an English character string when inputting English text, or may input a Japanese character string by inputting Roman characters (step S1).
- the user does not need to make the system 1 recognize in advance the language of the character string that he / she will input.
- the user may input “okuru” and Japanese, or “e-mail” and English Good.
- the input support system 1 executes two types of processing using the character string.
- One is a process in which the input character string is regarded as English and a word or phrase including the character string is searched from the database (step S2).
- the other is processing that regards the input character string as Japanese and acquires an English translation related to the character string from the database (step S3).
- the input support system 1 presents the results obtained by these two processes to the user at a time as input candidates (step S4).
- English phrases are obtained, so all the phrases presented to the user as input candidates are in English.
- the phrase in this specification means one or both of a word and a phrase.
- the input support system 1 includes a user terminal 10, an input support server (input support device) 20, and a language database (storage unit) 30.
- the user terminal 10 and the input support server 20 can communicate bidirectionally via a communication network such as the Internet or a dedicated line. Further, the input support server 20 can access the language database 30 via a similar communication network.
- the user terminal 10 is a computer used by a user who wants to create an English sentence.
- the type of the user terminal 10 is not limited, and may be, for example, a stationary or portable personal computer, or a portable terminal such as a high-function mobile phone (smart phone), a mobile phone, or a personal digital assistant (PDA).
- PDA personal digital assistant
- the number of user terminals 10 present in the input support system 1 is not limited.
- the user terminal 10 includes an editor 11 that executes character input and display.
- the editor 11 has a function of communicating with the input support server 20, but the specific method for mounting the editor 11 is not limited.
- the editor 11 may be implemented on a web browser (Web Browser) or may be incorporated in a specific word processor application program.
- the input support server 20 is a computer that searches for input candidates based on the input character string and presents them to the user when Japanese characters in Roman letters or English words are input.
- the input support server 20 refers to the language database 30 in order to execute such processing.
- the language database 30 is a device that stores parallel translations between English and Japanese.
- the specific configuration (mode) of the language database 30 and the data stored therein is not limited as long as the correspondence between the English phrase and the Japanese phrase expressed in Roman letters can be specified.
- an existing dictionary database or corpus may be used as the language database 30.
- Japanese in the language database 30 needs to be written in at least Roman characters.
- the language database 30 includes a bilingual database of English and Japanese written in Hiragana, Katakana, or Kanji (Japanese written in Kana), and Japanese written in Roman and Kanji. It may consist of a conversion table indicating correspondence with words.
- the editor 11 displays the input candidate English phrases separately for each Japanese meaning, so information relating to the Japanese meaning (for example, case frame dictionary information) is added to each phrase in advance. Yes.
- the language database 30 may include not only parallel translation data but also data indicating a list of English words / phrases. As shown in FIG. 1, since the input support system 1 executes two types of searches, the database or storage area to be referenced may be changed for each search process. For example, the input support system 1 uses a database in which only English expressions are registered when performing a search from English, and bilingual translation between two languages is registered when performing a search from Japanese. Existing databases may be used.
- the input support server 20 includes a CPU 201 that executes an operating system, application programs, and the like, a main storage unit 202 composed of ROM and RAM, and an auxiliary storage unit 203 composed of a hard disk and the like. And a communication control unit 204 composed of a network card, an input device 205 such as a keyboard and a mouse, and an output device 206 such as a display.
- Each functional component of the input support server 20 to be described later reads predetermined software on the CPU 201 or the main storage unit 202, and controls the communication control unit 204, the input device 205, the output device 206, and the like under the control of the CPU 201.
- the operation is realized by reading and writing data in the main storage unit 202 or the auxiliary storage unit 203. Data and a database necessary for processing are stored in the main storage unit 202 or the auxiliary storage unit 203.
- FIG. 3 shows that the input support server 20 is configured by a single computer, the functions of the input support server 20 may be distributed to a plurality of computers.
- the input support server 20 includes a receiving unit 21, a search unit 22, and a providing unit (output unit) 23 as functional components.
- the reception unit 21 is a functional element that receives an input character string in alphabet.
- the editor 11 of the user terminal 10 transmits the input character string to the input support server 20, and the reception unit 21 receives the input character string and outputs the character string to the search unit 22.
- the timing at which the editor 11 transmits the input character string to the input support server 20 is not limited.
- the editor 11 may transmit the character string to the input support server 20 for the first time when receiving an input of a character string having a predetermined length or more (for example, a character string having two or more characters).
- a character string having two or more characters By setting the lower limit of the length of the input character string received by the receiving unit 21, the editor 11 can be reduced by reducing the search load of words or phrases, or by suppressing the number of input candidates displayed on the user terminal 10. It can be made easier to use.
- the editor 11 may transmit an input character string when a delimiter (space) is input. By using one word as an input character string, it is possible to further reduce the search load and further reduce the number of input candidates.
- the editor 11 may transmit the input character string when the key input is interrupted for a predetermined time (for example, 2 seconds).
- a predetermined time for example, 2 seconds
- the editor 11 may transmit the input character string in response to the user pressing a search button, a transmission button, or the like. According to these two methods, an input character string made up of a plurality of words separated by spaces can be transmitted to the input support server 20.
- the receiving unit 21 or the editor 11 may automatically specify a plurality of words from an input character string without a delimiter. For example, when the receiving unit 21 or the editor 11 receives a character string “nimotsukeukuru” (corresponding to “send a package” in English), the receiving unit 21 or the editor 11 considers this character string as Japanese in Roman characters and performs morphological analysis. , “Nimoto”, “wo”, and “okuru” may be specified, and these words may be output to the search unit 22. In such word division, the reception unit 21 or the editor 11 deletes specific words (for example, particles) that are less necessary for the search, and outputs only the remaining words (for example, nouns and verbs) to the search unit 22. May be. For example, the reception unit 21 or the editor 11 may delete the Japanese particle “wo” and output only the noun “nimoto” and the verb “okuru” to the search unit 22.
- specific words for example, particles
- the reception unit 21 or the editor 11 may delete the
- the search unit 22 is a functional element that acquires input candidates corresponding to the input character string from the language database 30.
- the search unit 22 acquires an English word / phrase by two methods as shown in FIG.
- the search unit 22 acquires an English (first language) phrase including an input character string from the language database 30 as a first search result.
- the search unit 22 performs a search assuming that the input language is English. Therefore, when Japanese in Roman letters is input, the search unit 22 cannot acquire an English word / phrase in the first process (search from English). Of course, even if the user intends to input Japanese and there is an accidental occurrence of English including the character string, the search unit 22 acquires one or more English words / phrases in this first process.
- the search unit 22 searches for words such as “forget”, “forgot”, “forgive”, and phrases such as “get forgotten”, “get forgetful”, “forgive A for B”. get.
- the search unit 22 acquires an English (first language) phrase corresponding to a Japanese (second language) phrase including the input character string as a second search result.
- the search unit 22 performs a search assuming that the input language is Japanese. Therefore, when English is input, the search unit 22 performs the second process (from Japanese). Search) cannot get English phrases. Of course, even if the user intends to input English, if the Japanese language indicated by the character string happens to exist, the search unit 22 acquires one or more English words / phrases in the second process.
- the search unit 22 uses words such as “send”, “give”, and “lead” that can correspond to Japanese, and “send e-mail” that can correspond to this Japanese. to ",” e-mail the address of "," lead a gracieous life "and other phrases. Since one Japanese word may correspond to a plurality of English words, English phrases having various meanings may be extracted from one Japanese word as in this example.
- search unit 22 extracts a related word “kamoto (cargo)” in Japanese “nimoto” and a related word “soufu (send)” in Japanese “okuru” from the thesaurus, and uses these two related words.
- An English phrase may be acquired from the language database 30.
- search methods as described above, and the search unit 22 may search for a phrase using any one or more methods.
- the search unit 22 outputs the first and second search results obtained by such two types of processing to the providing unit 23. At this time, the search unit 22 determines whether each of the acquired English words / phrases has been extracted by a search from English (first processing) or a search from Japanese (first A flag indicating whether it has been extracted by the process 2) is added. Of course, both the first and second search results may be empty (NULL) due to a typographical error of the user.
- the search unit 22 When the input character string includes a plurality of words, the search unit 22 performs the above-described two types of search for each word, and finally acquires a common phrase extracted in all the input words. Good (search for common phrases). By searching for input candidates using all the words constituting the input character string, it is possible to present with high accuracy the word or phrase that the user intends to input.
- the input character string may consist only of English, may consist only of Japanese, or may be a mixture of Japanese and English.
- search unit 22 uses the phrase “send a package” (meaning “send a package” in Japanese) in both cases of Japanese “nimoto” and English “send”. Search for.
- the search unit 22 may perform the above two types of searches only for some of the words (search using some words). For example, the search unit 22 may perform the above two types of searches only for the last word in the input character string. By searching for input candidates using only some words, it is possible to avoid a situation in which too many or too few English phrases are extracted.
- the search unit 22 may perform both a search for a common phrase and a search using some words, and obtain these two types of search results.
- the search unit 22 uses a flag for distinguishing whether each word obtained as an input candidate is a word obtained by searching for a common phrase or a word obtained by a search using some words. Is added.
- the providing unit 23 is a functional element that transmits the search result by the search unit 22 to the user terminal 10 as an input candidate.
- each English word / phrase included in the search result is associated with a flag indicating what kind of search the word / phrase is extracted.
- the editor 11 receives the input candidates and displays them on the screen.
- the editor 11 obtains a character string having a predetermined length or more input in alphabet (step S11), and transmits the input character string to the input support server 20 (step S12).
- the reception unit 21 receives the input character string (step S12, reception step), and then the search unit 22 searches for input candidates based on the character string. Specifically, the search unit 22 searches for an English phrase including the input character string (step S13, search step), and searches for an English phrase corresponding to the Japanese phrase including the input character string (step S13). S14, search step). Note that the search unit 22 may execute a search from English (step S13) after performing a search from Japanese (step S14). Finally, the providing unit 23 transmits the results of these two types of searches as input candidates to the user terminal 10 (step S15, output step).
- the editor 11 receives the input candidates and displays them on the screen (step S16, output step).
- the user can select one of the displayed input candidates, and the selected English phrase is displayed on the editor 11 as a fixed phrase.
- the editor 11 transmits the input character string “okuru” to the input support server 20, and the search unit 22 executes two types of searches based on the character string. Since there is no English word including “okuru”, the search unit 22 cannot obtain an English phrase by searching from English. However, since “okuru” in Japanese has a meaning equivalent to several words such as “send”, “give”, “present”, and “lead” in English, the search unit 22 performs a search from Japanese. Get an English phrase by running. Therefore, the providing unit 23 transmits one or a plurality of English words as input candidates to the user terminal 10, and the editor 11 displays the input candidates on the screen.
- FIG. 10 An example of input candidates displayed on the user terminal 10 is shown in FIG.
- the editor 11 displays a list of input candidates in the candidate field C.
- Each of the searched English phrases E is accompanied by a Japanese translation J.
- the title “word (from Japanese)” in the upper left of the candidate column indicates that the current word / phrase is obtained by searching from Japanese.
- the editor 11 can determine the title by referring to the flag associated with the acquired word / phrase.
- the editor 11 displays the input candidates separately for the word list L1 displayed in the leftmost column and the phrase list displayed in the other columns. Furthermore, the editor 11 divides the phrases into columns L2 to L4 for each meaning of Japanese “okuru”. In the example of FIG. 6, since the number of English words is larger than the height of the candidate field C, a scroll bar is displayed in the column L1, but the editor 11 displays this scroll bar as the number of words in each column. Or set to each column according to the number of phrases. Therefore, the user can look around the words in a specific column while continuing to look at the words in other columns. If the number of groups when the English phrase is divided into Japanese meanings is larger than the width of the candidate field C, the editor 11 sets a horizontal scroll bar for looking over the phrase column.
- the user can easily find and select a desired word or phrase.
- the editor 11 may display the acquired words or phrases without dividing them into words and phrases, but only after dividing them into meanings. Also in this case, the user can easily find and select a desired word / phrase.
- FIGS. 6 and 7 another meaning of the selected word may be indicated by a box B.
- the editor 11 transmits the input character string “e-mail” to the input support server 20, and the search unit 22 executes two types of searches based on the character string. Since there is no Japanese written in Roman letters as “e-mail”, the search unit 22 cannot obtain an English word / phrase when searching from Japanese. However, as a matter of course, the search unit 22 can acquire an English phrase including “e-mail” by executing a search from English. Even if the input character string is “e-mai”, the same result can be obtained. Therefore, the providing unit 23 transmits one or a plurality of English words as input candidates to the user terminal 10, and the editor 11 displays the input candidates on the screen.
- FIG. 12 shows an example of input candidates displayed on the user terminal 10 in this case.
- the editor 11 also writes a Japanese translation for each English phrase in the candidate field C.
- “e-mail” having a plurality of meanings is focused, and explanations of other meanings are displayed in box B in response to this focus setting.
- whether or not to implement the function of the box B in the editor 11 may be arbitrarily determined.
- the editor 11 displays the selected word / phrase in place of the input character string “e-mail”, thereby confirming the input.
- FIG. 13 shows an example of input candidates when the input character string includes a plurality of words.
- a list of English words corresponding to the last Japanese word “okuru” in the input character string is displayed.
- a list of English phrases including translations of both two Japanese “nimoto” and “okuru” included in the input character string is displayed.
- columns L3 and L4 a list of English phrases including the translation of the last Japanese word “okuru” in the input character string is displayed.
- the editor 11 can determine in which column the phrase should be displayed by referring to the flag associated with each phrase.
- the editor 11 selects only the last word “okuru” in the input character string. Replace the selected word and confirm the input. In other words, the editor 11 replaces only the word used as the search condition for the English phrase in the input support server 20 with the phrase specified by the user. In this case, the user can know which word has been replaced.
- the editor 11 confirms the input by replacing the two Japanese words “nimoto” and “okuru” included in the input character string with the selected phrase. Let Therefore, the user can immediately start the next editing work.
- the editor 11 can determine the replacement portion of the input character string by referring to the flag of each word in the input candidate.
- the input character string includes two words. However, as described above, the input character string may include three or more words. For example, the input character string may include three words “Japanese” “okuru”, “nimoto”, and “kuukubin (corresponding to“ airmail ”in English)”.
- the input support program P includes a client program P1 for causing the user terminal 10 to function the editor 11 and a server program P2 for causing the computer to function as the input support server 20.
- the client program P1 includes an editor module P11, and the function of the editor 11 is realized by this module.
- the server program P2 includes a main module P20, a reception module P21, a search module P22, and a provision module P23.
- the main module P20 is a part that controls the functions of the input support server 20 in an integrated manner.
- the functions realized by executing the reception module P21, the search module P22, and the provision module P23 are the same as the functions of the reception unit 21, the search unit 22, and the provision unit 23, respectively.
- Each of the client program P1 and the server program P2 is provided after being fixedly recorded on a tangible recording medium such as a CD-ROM, DVD-ROM, or semiconductor memory.
- Each program P1, P2 may be provided via a communication network as a data signal superimposed on a carrier wave.
- the process of searching for a phrase from the input character string itself without searching (search from English) and the process of translation based on the input character string (search from Japanese) ) are executed, and English words and phrases obtained from these two types of processing are collectively output as input candidates.
- the user needs to input a sentence or a sentence after performing an operation such as switching an input mode.
- the user since two systems of processing are performed on one input character string and both search results are output, the user does not specify the language of the input character string himself (for example, input) You can enter words in the desired language (without switching modes).
- This system 2 is the first embodiment (input support system) in that when the user selects one word from input candidates and confirms the input, the word corresponding to the slot of the word is further presented. Different from 1).
- the input support system 2 includes an input support server 20A.
- “Slot” in this specification is an undetermined part set in one word selected from input candidates, and an object, a complement or the like can be inserted into the undetermined part. This slot is set before or after a word, or before, after or in the middle of a phrase. For example, “someone” in the phrase “give someone a birthday present” is a slot set in the phrase. As another example, a portion following the phrase “e-mail the address of” can also be set as a slot. Alternatively, “A” and “B” in the phrase “provide A with B” are also examples of slots. If words and phrases that apply to such a slot are further presented, it can be expected that the text input is further simplified.
- the input support server 20A of the present embodiment has a slot processing unit (additional processing unit) in addition to the three functional components in the first embodiment in order to realize the additional function as described above. 24.
- the editor 11 further has a function for obtaining a word that may be applied to the slot.
- information indicating the slot position is associated in advance with at least some of the English phrases that can be displayed as input candidates on the user terminal 10.
- the editor 11 When the user selects one word from the input candidates, the editor 11 confirms the input of the word and transmits information on the word to the input support server 20A. The editor 11 sets the focus at the slot position. If there are a plurality of slots, the editor 11 sets the focus to one of the slots (for example, the first slot). Thereafter, the editor 11 receives a further input candidate sent from the input support server 20 in response to the transmission of the fixed phrase information, and displays the input candidate on the screen. This further input candidate is a phrase that can be inserted into the slot.
- the slot processing unit 24 is a functional element that, when input of an English word / phrase having a slot is confirmed, acquires an English word / phrase inserted into the slot and outputs the acquired word / phrase as a further input candidate.
- the slot processing unit 24 acquires a phrase that may be inserted into the slot based on the fixed phrase information received from the editor 11, and transmits the processing result to the user terminal 10.
- the method of searching for a phrase that applies to a slot is not limited.
- the slot processing unit 24 determines whether Phrases that can be inserted into the slots may be extracted from the language database 30.
- the slot processing unit 24 may determine a phrase that can be inserted into a slot of a definite phrase by using a conventional natural language processing technique. For example, the slot processing unit 24 obtains further input candidates by using an n-gram method for English sentences, and further inputs by referring to the parallel translation data including information of the case frame dictionary. Candidates may be obtained. As described above, the slot processing unit 24 uses both the search from English and the search from Japanese as in the case of obtaining the input candidate (first input candidate) described in the first embodiment. Candidates can be obtained. In this case, the slot processing unit 24 associates a flag indicating which of the two types of search processing the obtained word / phrase is obtained with the word / phrase. Note that the natural language processing used in this processing is not limited to n-grams and case frame dictionaries, and any method can be adopted.
- the slot processing unit 24 may extract further input candidates by searching an English corpus in the language database 30. Specifically, the slot processing unit 24 extracts a sentence including a definite word / phrase, analyzes one or more extracted sentences, and specifies a word / phrase actually inserted at the slot position together with the number of appearances. Then, the slot processing unit 24 creates a list by sorting the specified one or more words in descending order of the number of appearances, and sets this list as a further input candidate. For example, the slot processing unit 24 sets the word groups “you”, “her”, “him”, and “Tom” as further input candidates for the fixed phrase “give someone a birthday present” in descending order of the number of appearances.
- the slot processing unit 24 may use PMI (Pointwise Mutual Information) or the like to set a characteristic word that is not high in frequency but is a subsequent input candidate.
- PMI Pointwise Mutual Information
- the slot processing unit 24 analyzes a sentence or a sentence located before the current fixed phrase on the editor 11 to identify a phrase that has a high probability of being inserted into the slot. May be. For example, it is assumed that a sentence “His birthday is in December. I will” has already been input. In this case, when the user inputs “okuru” and selects the phrase “give someone a birthday present” from the input candidates displayed according to the operation, the slot processing unit 24 uses the word “His” as a clue. “Him” is extracted as a further input candidate. In the fourth method, the slot processing unit 24 needs to receive not only the information of the definite word / phrase but also the information of the sentence or sentence ahead of it from the editor 11. According to the fourth method, it is possible to present in advance a phrase that a user who is inputting a sentence or sentence is likely to input.
- the slot processing unit 24 may display the word derived according to the fourth technique out of the word group extracted according to the third technique in preference to the word not derived according to the technique (first). Method 5).
- the slot processing unit 24 may set the display order of each word in the word group so that the display order of the words derived by the fourth technique is earlier than the display order of other words.
- the method of acquiring words that can be inserted into the slot there is no limitation on the method of acquiring words that can be inserted into the slot. Further, the display order of the words is not limited to the above.
- the slot processing unit 24 may arrange the acquired words in alphabetical order.
- the slot processing unit 24 may extract further input candidates for each slot by an arbitrary method (for example, any one of the first to fifth methods described above).
- the slot processing unit 24 may transmit information on a natural combination of words inserted in each of at least two slots of the plurality of slots to the editor 11 together with words that can be inserted in each slot.
- the combination information may be stored in the language database 30 or may be held in advance by the slot processing unit 24.
- the slot processing unit 24 combines the combination information Ca indicating that the combination of “library” and “borrow” is more natural than the combination of “library” and “rent”, and “rental shop” and “ The combination information Cb indicating that the combination of “rent” is more natural than the combination of “rental shop” and “borrow” is read from the language database 30 and transmitted to the editor 11.
- steps S11 to S16 are the same as that in the first embodiment. Thereafter, when the user selects one word / phrase from the input candidates and confirms the input (step S17), the editor 11 sets the focus to the slot of the confirmed word / phrase (step S18).
- the editor 11 sets the focus on the entire word “someone”.
- the editor 11 sets the focus after “of”.
- the user can immediately start the next editing operation without performing an operation such as moving the cursor. For example, by setting the focus to the entire word, the user can immediately overwrite the word by selecting from the input candidates or by directly inputting the word.
- the editor 11 transmits information on the fixed phrase to the input support server 20 (step S19).
- the slot processing unit 24 receives the information (step S19), and acquires an English phrase that can be inserted into the slot by any method described above (step S20). Then, the slot processing unit 24 transmits the extracted word / phrase as a further input candidate to the user terminal 10 (step S21).
- the editor 11 receives the further input candidates and displays them on the screen (step S22).
- the selected English word is inserted into the slot, thereby confirming the input.
- the editor 11 may determine the words of each slot one by one based on the user's selection operation. Alternatively, the editor 11 provides the user with an opportunity to reselect the words and phrases in each slot by confirming the words and phrases in all slots at once after the user performs a selection operation on all slots. Also good.
- the editor 11 When the combination information is received, the editor 11 automatically selects and presents the words / phrases of the other slot more naturally corresponding to the word / phrase selected by the user in one slot based on the information, or A more natural phrase may be presented to the user as help information without being automatically selected. In this way, by presenting a natural combination of phrases, the user can easily create a natural sentence or sentence.
- the editor 11 has two English words “rent” and “borrow” that are candidates for the first slot, English words that are candidates for the second slot, “library” and “rental shop”, and the combination information described above. Assume that Ca and Cb are received. Then, it is assumed that the user selects “library” in the second slot after selecting “rent” in the first slot. In this case, the editor 11 replaces “rent” in the first slot with “borrow” or displays help information indicating that “borrow” is a more natural word according to the user operation. .
- the input is confirmed by selecting the English phrase “chuck a ball to” from the input candidates.
- the editor 11 transmits information on this phrase to the input support server 20.
- the slot processing unit 24 acquires further input candidates based on this phrase and transmits the candidates to the user terminal 10.
- the editor 11 displays the further input candidates on the screen as shown in FIG.
- the editor 11 lists further input candidates in the candidate field C.
- further input candidates are displayed separately for a phrase (column L1) obtained by a search from English and a phrase (column L2) retrieved by a search from Japanese. Whether each word / phrase belongs to one of the two groups can be determined by referring to a flag associated with the word / phrase.
- the editor 11 may set the scroll bar in each column according to the number of words or phrases in each column.
- the input support program P for realizing the input support system 2 will be described with reference to FIG.
- the input support program P of this embodiment is different from that of the first embodiment in that the server program P2A includes a slot processing module P24 in addition to the main module P20, the reception module P21, the search module P22, and the provision module P23.
- the functions realized by executing the slot processing module P24 are the same as the functions of the slot processing unit 24 described above.
- the second embodiment described above has the same effect as the first embodiment.
- word candidates further input candidates
- word candidates that can be inserted into the slot are presented, so that the user can more easily read a sentence or a sentence. Can be created.
- a client / server type input support system is shown, but an input support apparatus having a function of the input support server 20 or 20A may be adopted by the user terminal or the editor.
- the editor 11 ⁇ / b> A mounted on the user terminal 10 ⁇ / b> A may include a reception unit 21, a search unit 22, a providing unit 23, and a slot processing unit 24.
- the search unit 22 and the slot processing unit 24 access the language database 30 via the network.
- the language database 30 may be provided in the input support server 20 or 20A. Similarly, in the modification of FIG. 19, the language database 30 may be provided in the user terminal 10A.
- the input support system accepts English input or Japanese input in Roman letters to support English input. That is, in the above embodiment, the first language is English and the second language is Japanese. However, which language is used as the first and second languages is a matter that can be arbitrarily determined. However, the first language is a language (for example, French or German) expressed using alphabets. On the other hand, the second language may be a language (for example, Chinese) that does not normally use an alphabet but can be expressed in alphabet, or a language (for example, French, German, or English) that is expressed by using an alphabet. Therefore, both the first and second languages may be Latin languages.
- the display method of input candidates and further input candidates on the editor is not limited.
- the editor 11 may display only one scroll bar for the entire candidate field C, or may display the input candidates in only two columns, a word column and a phrase column.
- a phrase that is not displayed may be displayed using a page switching technique instead of a scroll bar.
- the editor 11 sets the focus to the slot and then displays further input candidates.
- the input support system may end the process at the time when the focus is set to the slot, and may not acquire further input candidates.
Abstract
Description
図1~3を用いて、第1実施形態に係る入力支援システム(以下では単に「システム」ともいう)1の機能及び構成を説明する。このシステム1は英文入力を支援するコンピュータシステムである。このシステム1を用いることで、英語に不慣れなユーザも簡単に英文を作成することが期待できる。
次に、第2実施形態に係る入力支援システム2を説明する。このシステム2は、ユーザが入力候補から一の語句を選択して入力を確定させた際にその語句のスロット(slot)に当てはまる語句を更に提示するという点で、第1実施形態(入力支援システム1)と異なる。このために、入力支援システム2は入力支援サーバ20Aを備えている。
Claims (20)
- ローマ字による入力文字列を受け付ける受付部と、
アルファベットで表記される第1の言語の語句と、該第1の言語の語句に対応する第2の言語の語句のローマ字表記と、をそれらの対応関係が特定できる態様で記憶する記憶部から、前記入力文字列を含む前記第1の言語の語句を検索する第1の処理と、該記憶部から、該入力文字列を含む前記ローマ字表記に対応する前記第1の言語の語句を検索する第2の処理とを実行する検索部と、
前記第1の処理の結果と前記第2の処理の結果とを入力候補として出力する出力部と
を備える入力支援装置。 - 前記入力文字列が複数の単語を含み、
前記検索部が、前記複数の単語のそれぞれについて前記第1の処理及び前記第2の処理を実行し、
前記出力部が、前記複数の単語のすべてにおいて検索された共通の前記第1の言語の語句を前記入力候補として出力する、
請求項1に記載の入力支援装置。 - ユーザにより選択された前記共通の第1の言語の語句が前記入力文字列の全体と置換されることで、該選択された語句の入力が確定する、
請求項2に記載の入力支援装置。 - 前記入力文字列が複数の単語を含み、
前記出力部が、前記複数の単語のうち一部の単語において検索された前記第1の言語の語句を前記入力候補として出力する、
請求項1~3のいずれか一項に記載の入力支援装置。 - 前記一部の単語が、前記入力文字列において最後に位置する単語である、
請求項4に記載の入力支援装置。 - ユーザにより選択された前記第1の言語の語句が、前記複数の単語のうち該選択された語句の検索条件として用いられた単語のみと置換されることで、該選択された語句の入力が確定する、
請求項4又は5に記載の入力支援装置。 - 前記入力候補が、未確定部分を示す情報が付加された前記第1の言語の語句を含み、
前記未確定部分を有する語句の入力が確定した場合に、該未確定部分にフォーカスが設定される、
請求項1~6のいずれか一項に記載の入力支援装置。 - 前記入力候補が、未確定部分を示す情報が付加された前記第1の言語の語句を含み、
前記未確定部分を有する語句の入力が確定した場合に、該未確定部分に挿入され得る語句を取得し、取得した語句を更なる入力候補として出力する追加処理部を更に備える、
請求項1~7のいずれか一項に記載の入力支援装置。 - 前記追加処理部が、語句の未確定部分と該未確定部分に挿入可能な語句との対応が記憶された所定の記憶部を参照することで、前記未確定部分に挿入され得る語句を取得する、
請求項8に記載の入力支援装置。 - 前記追加処理部が、前記第1の言語及び前記第2の言語のうち少なくとも一方の言語のコーパスが記憶された所定の記憶部を参照することで、前記未確定部分に挿入され得る語句を取得する、
請求項8又は9に記載の入力支援装置。 - 前記更なる入力候補が前記コーパス内での出現回数に従った順番で表示される、
請求項10に記載の入力支援装置。 - 前記追加処理部が、前記入力文字列より前に位置する確定済の文字列を解析することで、前記未確定部分に挿入され得る語句を推定する、
請求項8~11のいずれか一項に記載の入力支援装置。 - 前記更なる入力候補において、前記確定済の文字列の解析により推定された語句が、該解析により推定されなかった語句よりも優先して表示される、
請求項12に記載の入力支援装置。 - 前記入力候補に含まれる前記第1の言語の語句に前記未確定部分が複数存在する場合に、前記追加処理部が、各未確定部分について前記更なる入力候補を取得すると共に、第1の未確定部分に挿入され得る語句と第2の未確定部分に挿入され得る語句との自然な組合せを示す組合せ情報を取得し、該組合せ情報を各未確定部分の更なる入力候補と共に出力し、
前記第1及び第2の未確定部分のうちの一方の未確定部分におけるユーザ操作に応じて、他方の未確定部分に挿入され得る語句が前記組合せ情報に基づいて提示される、
請求項8~13のいずれか一項に記載の入力支援装置。 - 前記入力候補が単語の列とフレーズの列とに分かれて表示される、
請求項1~14のいずれか一項に記載の入力支援装置。 - 前記フレーズが語義に従って更に複数の列に分かれて表示される、
請求項15に記載の入力支援装置。 - 前記入力候補が語義に従って複数の列に分かれて表示される、
請求項1~14のいずれか一項に記載の入力支援装置。 - 各列が独立してスクロールするように前記入力候補が表示される、
請求項15~17のいずれか一項に記載の入力支援装置。 - 入力支援装置により実行される入力支援方法であって、
ローマ字による入力文字列を受け付ける受付ステップと、
アルファベットで表記される第1の言語の語句と、該第1の言語の語句に対応する第2の言語の語句のローマ字表記と、をそれらの対応関係が特定できる態様で記憶する記憶部から、前記入力文字列を含む前記第1の言語の語句を検索する第1の処理と、該記憶部から、該入力文字列を含む前記ローマ字表記に対応する前記第1の言語の語句を検索する第2の処理とを実行する検索ステップと、
前記第1の処理の結果と前記第2の処理の結果とを入力候補として出力する出力ステップと
を含む入力支援方法。 - ローマ字による入力文字列を受け付ける受付部と、
アルファベットで表記される第1の言語の語句と、該第1の言語の語句に対応する第2の言語の語句のローマ字表記と、をそれらの対応関係が特定できる態様で記憶する記憶部から、前記入力文字列を含む前記第1の言語の語句を検索する第1の処理と、該記憶部から、該入力文字列を含む前記ローマ字表記に対応する前記第1の言語の語句を検索する第2の処理とを実行する検索部と、
前記第1の処理の結果と前記第2の処理の結果とを入力候補として出力する出力部と
をコンピュータに実行させる入力支援プログラム。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/650,220 US20150309991A1 (en) | 2012-12-06 | 2013-08-09 | Input support device, input support method, and input support program |
JP2014526291A JP5897711B2 (ja) | 2012-12-06 | 2013-08-09 | 入力支援装置、入力支援方法、及び入力支援プログラム |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US201261734026P | 2012-12-06 | 2012-12-06 | |
US61/734026 | 2012-12-06 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014087704A1 true WO2014087704A1 (ja) | 2014-06-12 |
Family
ID=50883135
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2013/071712 WO2014087704A1 (ja) | 2012-12-06 | 2013-08-09 | 入力支援装置、入力支援方法、及び入力支援プログラム |
Country Status (3)
Country | Link |
---|---|
US (1) | US20150309991A1 (ja) |
JP (1) | JP5897711B2 (ja) |
WO (1) | WO2014087704A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5728630B1 (ja) * | 2014-07-31 | 2015-06-03 | 楽天株式会社 | 情報処理装置、情報処理方法、プログラム、記憶媒体 |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH117448A (ja) * | 1997-06-17 | 1999-01-12 | Omron Corp | 記録媒体および文字入力装置 |
JP2001265765A (ja) * | 2000-03-23 | 2001-09-28 | Seiko Instruments Inc | 電子辞書 |
JP2009059139A (ja) * | 2007-08-31 | 2009-03-19 | Seiko Instruments Inc | 辞書検索装置、辞書検索方法及び辞書検索プログラム |
Family Cites Families (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11338858A (ja) * | 1998-05-22 | 1999-12-10 | Toshiba Corp | 入力予測装置、入力予測方法及び入力予測プログラムを記録した記録媒体 |
US8137105B2 (en) * | 2003-07-31 | 2012-03-20 | International Business Machines Corporation | Chinese/English vocabulary learning tool |
US7376648B2 (en) * | 2004-10-20 | 2008-05-20 | Oracle International Corporation | Computer-implemented methods and systems for entering and searching for non-Roman-alphabet characters and related search systems |
CN101286094A (zh) * | 2007-04-10 | 2008-10-15 | 谷歌股份有限公司 | 多模式输入法编辑器 |
US20100121870A1 (en) * | 2008-07-03 | 2010-05-13 | Erland Unruh | Methods and systems for processing complex language text, such as japanese text, on a mobile device |
JP5320925B2 (ja) * | 2008-09-18 | 2013-10-23 | 富士通株式会社 | 情報検索方法、情報検索装置及び情報検索プログラム |
US8635539B2 (en) * | 2008-10-31 | 2014-01-21 | Microsoft Corporation | Web-based language translation memory compilation and application |
EP2680159B1 (en) * | 2010-07-13 | 2020-01-15 | Motionpoint Corporation | Dynamic language translation of a message |
US10133737B2 (en) * | 2010-08-26 | 2018-11-20 | Google Llc | Conversion of input text strings |
JP5739718B2 (ja) * | 2011-04-19 | 2015-06-24 | 本田技研工業株式会社 | 対話装置 |
US8706472B2 (en) * | 2011-08-11 | 2014-04-22 | Apple Inc. | Method for disambiguating multiple readings in language conversion |
US8775165B1 (en) * | 2012-03-06 | 2014-07-08 | Google Inc. | Personalized transliteration interface |
US20130326347A1 (en) * | 2012-05-31 | 2013-12-05 | Microsoft Corporation | Application language libraries for managing computing environment languages |
-
2013
- 2013-08-09 WO PCT/JP2013/071712 patent/WO2014087704A1/ja active Application Filing
- 2013-08-09 US US14/650,220 patent/US20150309991A1/en not_active Abandoned
- 2013-08-09 JP JP2014526291A patent/JP5897711B2/ja active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH117448A (ja) * | 1997-06-17 | 1999-01-12 | Omron Corp | 記録媒体および文字入力装置 |
JP2001265765A (ja) * | 2000-03-23 | 2001-09-28 | Seiko Instruments Inc | 電子辞書 |
JP2009059139A (ja) * | 2007-08-31 | 2009-03-19 | Seiko Instruments Inc | 辞書検索装置、辞書検索方法及び辞書検索プログラム |
Non-Patent Citations (2)
Title |
---|
FUMIHIKO SHIBATA: "Switch shitemo Smooth ni Sosa dekiru Windows no Omotenashi Dai 9 Kai Kaiteki na Bunsho Sakusei Kankyo no Jitsugen", MAC PEOPLE, vol. 17, no. 15, 29 August 2011 (2011-08-29), pages 116 - 119 * |
MIE HIRAYAMA ET AL.: "Tango Toroku de Manzoku... dewa Mottainai!! IME Moji Nyuryoku Shoryakuka no Ogi", PC MODE, vol. 9, no. 11, 24 November 2004 (2004-11-24), pages 103 - 115 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5728630B1 (ja) * | 2014-07-31 | 2015-06-03 | 楽天株式会社 | 情報処理装置、情報処理方法、プログラム、記憶媒体 |
WO2016016994A1 (ja) * | 2014-07-31 | 2016-02-04 | 楽天株式会社 | 情報処理装置、情報処理方法、プログラム、記憶媒体 |
Also Published As
Publication number | Publication date |
---|---|
JP5897711B2 (ja) | 2016-03-30 |
US20150309991A1 (en) | 2015-10-29 |
JPWO2014087704A1 (ja) | 2017-01-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5761833B2 (ja) | ユーザ部分入力のための辞書候補 | |
US7774193B2 (en) | Proofing of word collocation errors based on a comparison with collocations in a corpus | |
KR100641838B1 (ko) | 심벌문자를 이용한 다국언어 문자 입력 방법 및 문자 입력시스템 | |
US20060020576A1 (en) | Search system reusing search condition and the related method | |
JP2003223437A (ja) | 正解語の候補の表示方法、スペルチェック方法、コンピュータ装置、プログラム | |
US8782049B2 (en) | Keyword presenting device | |
JP4502615B2 (ja) | 類似文検索装置、類似文検索方法、およびプログラム | |
RU2595531C2 (ru) | Способ и система генерирования определения слова на основе множественных источников | |
JP2009059140A (ja) | 電子辞書、電子辞書の検索方法及び電子辞書の検索プログラム | |
JP5897711B2 (ja) | 入力支援装置、入力支援方法、及び入力支援プログラム | |
JP4491389B2 (ja) | 電子機器、プログラム、及びプログラムを記録した記録媒体 | |
JP5380989B2 (ja) | 辞書機能を備えた電子装置およびプログラム | |
JP5271526B2 (ja) | 商標検索システム及び商標検索サーバ | |
JP2008059169A (ja) | 中国語例文検索装置および中国語例文検索処理プログラム | |
JP2007257188A (ja) | 辞書検索装置およびその制御プログラム | |
JP2010146061A (ja) | 例文表示装置、例文表示方法および例文表示プログラム | |
JP6048601B2 (ja) | 情報表示装置、プログラムおよび単語情報表示方法 | |
JPS61248160A (ja) | 文書情報登録方式 | |
JP2007257189A (ja) | 辞書検索装置及びその制御プログラム | |
JP2008102802A (ja) | 文書支援システム及び文書支援プログラム | |
JP6551066B2 (ja) | 電子辞書 | |
JP2000148747A (ja) | 変換候補表示方法,およびこの方法によるかな漢字変換用のプログラムの記録媒体ならびにかな漢字変換装置 | |
JP2023114935A (ja) | 仮名漢字変換/kearm変換式日本語入力システム | |
JP5767784B2 (ja) | 情報処置装置、文字列候補確定方法及びプログラム | |
JP2009265845A (ja) | 例文検索装置および例文検索処理プログラム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
ENP | Entry into the national phase |
Ref document number: 2014526291 Country of ref document: JP Kind code of ref document: A |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13861281 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14650220 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 13861281 Country of ref document: EP Kind code of ref document: A1 |