Character input support device, character input support system and program

Abstract

To enable a user to generate a structured character string or the like without specific work by only creating an input character string and performing an operation to convert into a Kana and Kanji character string.SOLUTION: A character input support device according to the present invention comprises: an input unit in which an input character string is input; a knowledge base that returns one or more entities including attribute information corresponding to the input character string according to a search request using the input character string; a character input support unit in which the one or more entities returned by the knowledge base are input to output the entity, and that structures the input character string according to a user operation selecting any one of the output entities, and outputs the structured input character string; a preview unit that displays the entity output by the character input support unit; and a data storage unit that stores the structured input character string.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a character input assist device, a character input assist system, and a program for generating a structured character string from a character string input by a user. [Background technology]

[0002] Conventionally, efforts have been made to structure character strings containing various data by adding metadata using XML, RDF, or the like, to make them easier to reuse. Furthermore, a formatter that supports the generation of such structured documents has been proposed (Patent Document 1). Such structured character strings will be hereinafter referred to as "structured character strings," and documents that contain only "structured character strings" will be referred to as "structured documents."

[0003] Also, a technology has been proposed that supports the creation of structured documents and encourages their use by generating structured documents using a database (Patent Document 2).In addition, input support technologies and products have been proposed for supporting character input, such as IME (registered trademark), that convert a character string input using alphabets or kana (input character string) into a character string including kana and kanji (kana-kanji character string) (kana-kanji conversion) (Patent Document 3). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2002-312353 A [Patent Document 2] JP 2012-008663 A [Patent Document 3] JP 2014-029646 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, while conventional structured documents are structured as metadata for documents using XML and RDF, the data does not structure the text, for example, Japanese character strings themselves. Also, IME is a tool for inputting text and character strings themselves, and although there are examples of using a database to improve the accuracy of conversion, the text itself is a normal string of characters.

[0006] Once a normal character string becomes text, some of the information considered by the user when inputting it is lost. For example, when a user inputs "The weather in Shibuya today is sunny," information such as the fact that the user input "kyo" with the intention of reading it as "kyo" rather than "konnichi" and that the user input the whole area "Shibuya" in mind rather than pinpointing "Shibuya Station" is lost.

[0007] Additionally, while word breaks are obvious to the user, they are information that is lost when the text becomes a normal string. For example, unlike English, word breaks in Japanese are not obvious because they are made up of spaces, so word breaks are lost once the text becomes text. In addition, whether the user was consciously entering the characters "gravity acceleration" (the two consecutive words "gravity" and "acceleration"), or the single word "gravity acceleration" (gravity acceleration), may be obvious to the user, but is lost when the text becomes a normal string. It is difficult to infer in retrospect the information that is lost when the text becomes a normal string.

[0008] The present invention aims to enable a user to generate a structured character string and a structured document without any special work, simply by creating an input character string and converting it into a kana-kanji character string, and also to enable the data included in the input structured character string and structured document to be effectively and reliably used in other applications. [Means for solving the problem]

[0009] A character input assistance device according to the present invention includes an input unit into which an input string is input, a knowledge base that returns one or more entities including attribute information corresponding to the input string in response to a search request using the input string, a character input assistance unit that receives one or more entities returned by the knowledge base and outputs the entities, and that structures the input string in response to a user operation of selecting one of the output entities, and outputs the structured input string, a preview unit that displays the entities output by the character input assistance unit, and a data storage unit that stores the structured input string.

[0010] Furthermore, the attribute information includes information corresponding to the input string, and the knowledge base returns the entity including information corresponding to the input string used in the search request.

[0011] The character input support unit also sets the input entity as a conversion candidate for the input character string and outputs it to the preview unit.

[0012] Further, a user operation of selecting one of the output entities is an operation of identifying one of the conversion candidates for the input character string, and the preview unit receives the conversion candidates for the input character string from the character input assistance unit and displays them, and the character input assistance unit receives an operation of identifying one of the conversion candidates for the input character string via the input unit.

[0013] In addition, the character input assistance unit structures the input string by adding the entity returned by the knowledge base to the structured input string in response to an operation of identifying one of the conversion candidates for the input string.

[0014] Furthermore, a character input assistance system according to the present invention is a character input assistance system including: a knowledge base that returns an entity including attribute information corresponding to an input string in response to a search request using the input string; and a structuring processing device, wherein the structuring processing device includes an input unit that receives the input string; a character input assistance unit that receives one or more entities returned by the knowledge base in response to the search request using the input string and outputs the entities, and that structures the input string in response to a user operation of selecting one of the output entities and outputs the structured input string; a preview unit that displays the entities output by the character input assistance unit; and a data storage unit that saves the structured input string.

[0015] In addition, the program of the present invention causes a computer to execute the steps of: inputting an input string; returning one or more entities including attribute information corresponding to the input string in response to a search request using the input string; outputting the one or more entities returned in response to the search request, structuring the input string input in response to a user operation of selecting one of the output entities, and outputting the structured input string; displaying the output entities; and saving the structured input string. Effect of the Invention

[0016] According to the present invention, a user can generate a structured character string and a structured document without any special work by simply creating an input character string and converting it into a kana-kanji character string, and the data contained in the input structured character string and structured document can be used effectively and reliably in other applications. [Brief description of the drawings]

[0017] [Figure 1] FIG. 1 is a diagram showing a configuration of a character input assist device 1. As shown in FIG. [Figure 2A]FIG. 2A is a diagram illustrating entities contained in the knowledge base 12 shown in FIG. [Figure 2B] FIG. 2B is a diagram illustrating entities contained in the knowledge base 12 shown in FIG. [Diagram 3] FIG. 3 is a diagram illustrating an example of an entity that is displayed as a conversion candidate for the input character string "kyo" and selected by the user. [Figure 4] FIG. 4 is a diagram illustrating an example of a character string that the character input assist unit 102 adds to a structured character string or the like. [Diagram 5] FIG. 5 is a sequence diagram showing data input / output performed among the components of the character input assist device 1 shown in FIG. [Figure 6] FIG. 6 is a diagram illustrating an example of a structured character string generated by the character input assist device 1. As shown in FIG. [Figure 7A] FIG. 7A is a diagram illustrating an example of a UI screen used to correct a structured character string or the like. [Figure 7B] FIG. 7B is a diagram illustrating an example of a UI screen used to correct a structured character string or the like. [Figure 7C] FIG. 7C is a diagram illustrating an example of a UI screen used to correct a structured character string or the like. [Figure 7D] FIG. 7D is a diagram illustrating an example of a structured character string after correction. [Figure 7E] FIG. 7E is a diagram illustrating an example of an entity in the knowledge base 12 corresponding to "Ikebukuro (station name)" included in the corrected structured character string. [Figure 8] FIG. 8 is a sequence diagram showing input and output of data for generating and displaying subtitles. [Figure 9A] FIG. 9A is a diagram illustrating an example of a subtitle generated by the subtitle display device. [Figure 9B] FIG. 9B is a diagram illustrating an example of a subtitle generated by the subtitle display device. [Figure 10] FIG. 10 is a diagram showing the configuration of the manuscript system screen generating device 3. As shown in FIG. [Figure 11]FIG. 11 is a diagram illustrating an example of an original system screen generated by screen generating unit 30 of original system screen generating device 3 shown in FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] [Embodiment] First, a first embodiment of the present invention will be described in detail with reference to the accompanying drawings, in which the same components and processes are designated by the same reference numerals.

[0019] The character input assist device 1 will be described below. Fig. 1 is a diagram showing the configuration of the character input assist device 1. As shown in Fig. 1, the character input assist device 1 includes a structuring processing unit 10 and a knowledge base 12. The knowledge base 12 is also called a "knowledge database" or a "knowledge database".

[0020] The structuring processing unit 10 may be an information processing device (computer) equipped with a control unit (not shown) and a storage unit (not shown), as well as various input / output interfaces. Hereinafter, it will also be referred to as the structuring processing device 10. Similarly, the knowledge base 12 may be a database processing device (computer). In either case, the control unit realizes each function in this embodiment by appropriately reading and executing various programs stored in the storage unit. The control unit may be a CPU.

[0021] The structuring processing unit 10 includes an input unit 100, a character input assistance unit 102, a preview unit 104, and a data storage unit 106. The knowledge base 12 can be shared by a plurality of structuring processing devices 10. When a plurality of structuring processing devices 10 share the knowledge base 12, a character input assistance system can be formed.

[0022] The character input assist device 1 is an information processing device (computer) equipped with various input / output interfaces such as an input unit 100 and a preview unit 104, a display (not shown), and the like, in addition to the character input assist unit 102 and data storage unit 106 of the structuring processing device 10 and the knowledge base 12. The character input assist device 1 generates a structured character string and a structured document by structuring an input character string. Note that the structured document includes only the structured character string, and there is no substantial technical difference between the structured character string and the structured document.

[0023] The data storage unit 106 is a memory unit of the structuring processing device 10, and stores various programs and various data for causing the hardware group to function as the character input assist device 1. The data storage unit 106 may be a ROM, a RAM, a flash memory, a hard disk drive (HDD), or the like, and may be configured from multiple devices.

[0024] It is assumed that the character input assist device 1 is configured such that the structuring processing device 10 and the knowledge base 12 are integrated into one computer. However, the character input assist device 1 is not limited to this, and the structuring processing device 10 and the knowledge base 12 may be configured in separate computers.

[0025] First, the knowledge base 12 will be described. The knowledge base 12 is a database that stores and includes a plurality of entities that are prepared before the character input assist unit 102 converts a character string into a structured character string and a structured document. Each of the plurality of entities included in the knowledge base 12 is provided with a unique link.

[0026] Figures 2A and 2B are diagrams illustrating entities included in the knowledge base 12 shown in Figure 1. For convenience of illustration, the entities included in the knowledge base 12 are shown divided into two, Figures 2A and 2B.

[0027] 2A and 2B only exemplify 12 entities corresponding to seven words ("today," "of," "Shibuya," "of," "weather," "is," and "sunny") included in the structured string "The weather in Shibuya today is sunny." However, the entities included in knowledge base 12 are not limited to these. Knowledge base 12 may store and include numerous entities corresponding to all words that may be used to generate structured strings and structured documents.

[0028] 2A and 2B, words and their attributes are structured by being RDF-described in Turtle. In the following, a specific example will be given in which the entities of the knowledge base 12 and the structured character strings and structured documents generated by the structuring processing device 10 are RDF-described in Turtle. The structured character strings and structured documents do not necessarily need to be RDF-described, and may be described in other formats used for search, such as an RDB (Relational Data Base).

[0029] RDF is a framework that describes the relationship between a subject, predicate, and object using a set of three pieces of data called a triple, and expresses the relationship of information indicated by a Uniform Resource Identifier (URI) as a directed graph. SPARQL is a query language for extracting required information from data described in RDF by Turtle.

[0030] As shown in Figures 2A and 2B, knowledge base 12 includes 12 entities, including the links "<#Today>", "<#of>", "<#Shibuya (place name)>", "<#Shibuya1>", "<#Shibuya2>", "<#Weather>", "<#Weather1>", "<#Weather2>", "<#is>", "<#Sunny>", "<#Sunny1>" and "<#Sunny2>", respectively.

[0031] In an entity that includes the link "<#today>", attribute information of the word "today" included in the structured string or structured document is defined. This entity includes the attribute information "schema:name "today";", "schema:description "information about the date";", "schema:typicalAgeRange "8";" and "<reading> "kyo"" after the link "<#today>". When the attribute information "schema:typicalAgeRange" is not required, it may be indicated by an invalid value such as "schema:typicalAgeRange "Null";" or "schema:typicalAgeRange "-1";".

[0032] In the attribute information "schema:name "kyo";" included in this entity, "name" indicates a name, and this character string defines that the name of this entity is the kanji character string "kyo." Furthermore, this kanji character string "kyo" is used as a character for expressing this entity in a kana-kanji character string included in text, etc.

[0033] Among the attribute information "schema:description"information about date"" included in this entity, "description" indicates an array. In other words, this attribute information indicates that the word "today" indicated by the link "<#today>" is "information about a date."

[0034] The attribute information "schema:typicalAgeRange“8”;" included in this entity indicates, for example, that the age group that requires furigana for this string is under 8 years old. The attribute information "<reading>“kyo”" indicates the reading “kyo” of the word “kyo” included in the structured string or structured document.

[0035] As explained above, these character strings contained in this entity define the attribute information of "today" indicated by the link "<#today>", namely, the character string "today" when the word corresponding to this entity is displayed in text, the age group "under 8 years old" that requires furigana in the character string, and the reading of the word "today" "kyo".

[0036] In an entity that includes the link "<#no>", attribute information "schema:name"no"" of the two words "no" included in the above structured string or structured document is defined. Note that "Today's weather in Shibuya is sunny" includes two "no", but these two "no" are the same part of speech and are used in the same way, so the number of entities that include the link "<#no>" is counted as one.

[0037] In the entity that contains the link "<#Shibuya (place name)>", the attribute information of the word "Shibuya" contained in the above structured string or structured document, "schema:description"represents the place name of Shibuya"" and the " <nextpart><#Shibuya1>" is defined. The string " <nextpart>"<#Shibuya1>" indicates that the continuation of the attribute information of the word "Shibuya" is defined in the entity that includes the link "<#Shibuya1>".

[0038] In the entity that contains the link "<#Shibuya1>", the attribute information "schema:name"Shibu"", "<reading>"Shibu"", "schema:typicalAgeTange"12"" and " <nextpart>"<#Shibuya2>" is defined.

[0039] In the entity that contains the link "<#Shibuya2>", the attribute information "schema:name"Tani", "<reading>"ya", "schema:typicalAgeTange"8"" and " <nextpart><#Shibuya (place name)>" is defined. In this way, the attribute information " <nextpart>By using "<#Shibuya (place name)>", attribute information related to the word "Shibuya" can be used cyclically.

[0040] In the entity that contains the link "<#weather>", the attribute information "schema:name"weather"" of the word "Shibuya" contained in the above structured string or structured document and the " <nextpart>"<#Weather1>" is defined.

[0041] In the entity that contains the link "<#weather1>", the attribute information "<reading>"ten", "schema:typicalAgeTange"7"" and " <nextpart>"<#Weather2>" is defined.

[0042] In the entity that contains the link "<#weather2>", the attribute information "schema:name"ki"", "<reading>"ki"", "schema:typicalAgeTange"7"" and " <nextpart>"<#Weather>" is defined.

[0043] In an entity including the link "<# is>", attribute information "schema:name" is ";" is defined.

[0044] For entities that contain the link "<#Sunny>", the attribute information "schema:name"Sunny";" and " <nextpart><#Sunny1>" is defined.

[0045] For the entity containing the link "<#Sunny1>", the attribute information "schema:name"Sunny";", "<reading>"is", "schema:typicalAgeRange"7"" and " <nextpart><#Sunny2>" is defined.

[0046] For the entity that contains the link "<#Sunny2>", the attribute information "schema:name"れ";" and " <nextpart>"<#Sunny>" is defined.

[0047] It will be obvious to those skilled in the art that the kana-kanji string "Today's weather in Shibuya is sunny" can be obtained by arranging the characters indicated in the attribute information "schema:name" of the entity that contains the links corresponding to each word contained in the structured string or structured document shown in Figures 2A and 2B.

[0048] If required by the process of the character input assist device 1 or the user, each entity may further include attribute information such as parts of speech of words included in the structured character string and the structured document. Alternatively, if unnecessary for the character input assist device 1 or the user, the entity may not include the character string "schema:typicalAgeRange "7"" or the like in the attribute information.

[0049] The character input assistance unit 102 uses the character string input to the input unit 100 to make a search request for an entity in a query language such as SPARQL to the knowledge base 12. In response to the search request from the character input assistance unit 102, the character input assistance unit 102 searches for an entity corresponding to the character string used in the search request and an entity related to the character string.

[0050] Furthermore, the knowledge base 12 outputs the entity obtained as a result of the search to the character input assistance unit 102. Alternatively, the knowledge base 12 receives a search request from another device (not shown in FIG. 1) using a link, searches for an entity to which the link is attached in response to the received search request, and outputs the entity obtained as a search result to the character input assistance unit 102 that has output the search request or to another device.

[0051] 2A and 2B, knowledge base 12 generates a structured character string and a structured document by describing the character string indicating the information input from another device in a structured language as described with reference to Fig. 2A and Fig. 2B, and stores the structured character string and the structured document as a new entity. The new entity is stored in knowledge base 12 together with other entities that were previously included in knowledge base 12, and is included in knowledge base 12.

[0052] Next, each element of the structuring processing device 10 (FIG. 1) will be described. The input unit 100 includes input devices such as a keyboard, a microphone, and a mouse. An input character string is input to the input unit 100 by a user in response to an input operation on the input device by the user. This input character string is the subject of structuring in the structuring processing device 10. The input unit 100 also includes a voice recognition device (not shown) that converts the user's voice into a character string when an input character string is input from the microphone by the user's voice.

[0053] The input of the input character string using the microphone is performed, for example, when generating subtitles during the broadcast of a news program. The input unit 100 outputs an input signal indicating the input character string input by the user to the character input assistance unit 102.

[0054] The preview unit 104 displays on a display the input character string input via the input unit 100 and the character input support unit 102, as well as the conversion candidates, structured character strings, and structured documents input from the character input support unit 102, and presents them to the user.

[0055] As is generally known, a conversion candidate is a candidate kana-kanji character string obtained by converting an input character string of alphabetic characters or kana characters into kana-kanji. By displaying the conversion candidates and character strings by the preview unit 104, the user can check the input character string that he or she inputted into the structuring processing device 10, the conversion candidates for the input character string, and the structured character string and structured document generated in response to the selection of the conversion candidates.

[0056] The preview unit 104 may display the input character string, the conversion candidates, and the structured character string or the structured document in, for example, multiple windows displayed on the screen of a display. Alternatively, the preview unit 104 may display these in multiple areas obtained by dividing one window.

[0057] The data storage unit 106 stores and saves the structured character string and structured document input from the character input assistance unit 102, in addition to various programs and various data for causing the hardware group to function as the character input assistance device 1. The data storage unit 106 outputs the saved structured character string and structured document in response to a request from another device. The structured character string and structured document output from the data storage unit 106 to another device are provided to the requester and can be used for various purposes, as shown in the embodiments described below.

[0058] The character input assistance unit 102 is connected to the knowledge base 12, the input unit 100, the preview unit 104, and the data storage unit 106. The character input assistance unit 102 receives an input signal from the input unit 100. The character input assistance unit 102 converts the input signal into an input character string. Furthermore, the character input assistance unit 102 outputs an entity search request to the knowledge base 12 using the input character string, and receives one or more entities obtained as a search result by the knowledge base 12 that has received the search request.

[0059] The character input assistance unit 102 outputs the entity received from the knowledge base 12 to the preview unit 104 for display, and presents it to the user as a conversion candidate indicating a candidate for converting an input character into a kana-kanji character string. For example, when the character input assistance unit 102 receives a character string "kyo" or "k" from the input unit 100, the character input assistance unit 102 outputs a search request for an entity to the knowledge base 12 using the character string "kyo" or "k".

[0060] The knowledge base 12 searches for structured attribute information in each entity using a kana character string such as "kyo" used in the search request. The character input assistance unit 102 can also output a search request for an entity to the knowledge base 12 using the alphabet character "k". In this case, the character input assistance unit 102 can predict, from the input character string "k", one or more kana character strings beginning with a character in the "ka" row, such as "kyo (today)", "yinou (function)" and "kyonen (last year)", and set these as kana characters to be used in the search request in the knowledge base 12.

[0061] The knowledge base 12 outputs an entity including "<reading>"kyo"" (FIG. 2A) and, if necessary, a reading related to "kyo" or the like as attribute information to the character input assistance unit 102. The character input assistance unit 102 receives the entity output by the knowledge base 12, and outputs conversion candidates including the received entity to the preview unit 104 for display.

[0062] Fig. 3 is a diagram showing an example of an entity displayed as a conversion candidate for the input character string "kyo" and selected by a user. Note that the entity shown in Fig. 3 is the same as the entity containing the link "<#kyo>" shown in Fig. 2A. Fig. 4 is a diagram showing a link added to a structured character string or a structured document when the entity shown in Fig. 3 is selected by a user.

[0063] The user can, for example, use the mouse or the up, down, left and right buttons to select a desired entity from the conversion candidates displayed on the display by preview unit 104, and confirm the selection of the conversion candidate by pressing the enter key on the keyboard. When selection information indicating that the entity shown in Fig. 3 has been selected by the user is input via input unit 100, character input support unit 102 adds a link to this entity to the structured character string or structured document, as shown in Fig. 4.

[0064] When an entity shown in FIG. 3 is selected by a user, the character input assistance unit 102 adds a link of the selected entity to a structured character string or a structured document using the character strings "<#text1>", "schema;hasPart[", "schema:position 1;", and "<#term><#today>", as shown in FIG. 4.

[0065] When the link "<#term><#today>" is added to the beginning of a structured character string or a structured document, the character input assistance unit 102 adds the character string "@base<http: / / example.org> .", "@prefix schema:<http: / / schema.org> Add the link string "<#term><#today>." for this entity after ".", "<#text1>", "schema:hasPart[" and "schema:position 1;". In these strings, "position" indicates the position, "hasPart" indicates the part and "#term" indicates the word.

[0066] That is, the input unit 100, the character input assist unit 102, the preview unit 104, and the knowledge base 12 of the character input assist device 1 repeat the following processes (1) to (7) to generate text including a structured character string or a structured document. In detail, each time the user executes the process (1), the processes (2) to (7) are executed.

[0067] (1) A user inputs an input character string into the input unit 100; (2) The input unit 100 outputs the input character string to the character input support unit 102; (3) The character input support unit 102 outputs an entity search request to the knowledge base 12 using the input character string; (4) The knowledge base 12 receives a search request from the character input assistance unit 102 and outputs one or more entities corresponding to the input characters used in the search request as a search result to the character input assistance unit 102; (5) The character input assistance unit 102 receives entities obtained as search results from the knowledge base 12, and displays the received entities as conversion candidates on the preview unit 104 to present them to the user; (6) The character input assistance unit 102 receives, via the input unit 100, selection operation information indicating an operation by the user to select one of the conversion candidates; and (7) The character input assistance unit 102 adds a link of the selected entity to the structured character string or the structured document.

[0068] In this way, according to the character input assist device 1, the user can generate a structured kana-kanji character string or Japanese text simply by creating an input character string and converting it into a kana-kanji character string. In other words, these processes associate and connect the words included in the structured kana-kanji character string or Japanese text with the attribute information included in the entities of the knowledge base 12 via links.

[0069] Next, the overall processing of the character input assist device 1 will be described. Fig. 5 is a sequence diagram showing data input / output performed among the input unit 100, the character input assist unit 102, the preview unit 104, the data storage unit 106, and the knowledge base 12 of the character input assist device 1 shown in Fig. 1. As shown in Fig. 5, the user inputs an input character string to, for example, a keyboard (S100). The input unit 100 outputs an input signal indicating the input character string to the character input assist unit 102 (S102).

[0070] The character input assistance unit 102 receives an input signal from the input unit 100 and converts the input signal into an input string (S104). The character input assistance unit 102 uses the input string to make a search request for an entity to the knowledge base 12 (S106). The knowledge base 12 searches for an entity using the input string used in the search request, and obtains one or more entities as a search result (S108). The knowledge base 12 outputs the entity obtained as a result of the search to the character input assistance unit 102 (S110).

[0071] The character input assistance unit 102 outputs the entity as a conversion candidate to the preview unit 104 (S112), and the preview unit 104 presents it to the user (S114). When the user performs a selection operation to view the conversion candidates and select one of them (S116), the selection operation is input to the input unit 100, and selection information indicating the selection operation is output to the character input assistance unit 102 (S118).

[0072] The character input assistance unit 102 adds a link included in the entity (FIGS. 2A and 2B) selected as a conversion candidate to the structured character string or structured document (S120). The character input assistance unit 102 outputs the structured character string or structured document to which the link has been added to the data storage unit 106 for storage (S122), and outputs it to the preview unit 104 for display (S124). The preview unit 104 notifies the user that the conversion candidate selected by the user has been confirmed (S126). The processes of S100 to S126 are repeated to generate a structured character string or structured document.

[0073] Fig. 6 is a diagram illustrating a structured character string or a structured document generated by the character input assist device 1. The structuring processing device 10 of the character input assist device 1 uses the knowledge base 12 including the entities shown in Fig. 2A and Fig. 2B to repeat the processes described with reference to Fig. 3 to Fig. 5 as many times as necessary to generate a structured character string or a structured document as shown in Fig. 6.

[0074] The character input support unit 102 obtains links "<#today>" to "<#sunny>" in the entities corresponding to the words "today" to "sunny" included in the structured character string or structured document "The weather in Shibuya today is sunny." The character input support unit 102 sequentially obtains these links from the entities selected by the user from among the conversion candidates.

[0075] Furthermore, the character input assist unit 102 sequentially prefixes the character string "<#term>" to each of the acquired links "<#today>" to "<#sunny>". Furthermore, the character input assist unit 102 sequentially inserts the links "<#today>" to "<#sunny>" prefixed with "<#term>" into the "; ]" part of the character string "schema:hasPart[position 1; ]". As a result of these processes, the character input assist unit 102 generates the character string "schema:hasPart[position 1;<#term><#today>];" to "schema:hasPart[position 7;<#term><#sunny>].".

[0076] The character input support unit 102 adds a character string required for description in a structured character string or a structured document, for example, "@base<http: / / example.org> .", "@prefix schema:<http: / / schema.org> As a result of these processes, the character input assist unit 102 structures the kana-kanji character string "Today's weather in Shibuya is sunny" into a structured character string or a structured document.

[0077] Note that the process steps for generating a structured character string or a structured document from a character string shown here are merely examples, and character input assistance unit 102 may generate a structured character string or a structured document from a character string using a process step other than the above. In such a case, for example, character input assistance unit 102 first generates text "The weather in Shibuya today is sunny" from the input character string, divides this text into words using morphological analysis or the like, and further structures the words using links included in the entities obtained from knowledge base 12 to generate a structured character string or a structured document.

[0078] As described above, in a structured character string or a structured document, each of the words contained in the structured character string or the structured document includes a link to an entity contained in knowledge base 12. Therefore, by using the link included in each word to refer to the entity contained in knowledge base 12, the user can learn attributes such as the reading, meaning and concept of the kanji for each word contained in the structured character string or the structured document, and ambiguity is eliminated.

[0079] Furthermore, a structured character string and a structured document can be generated by the user simply performing normal text input operations using character input assist device 1. Therefore, the user does not need to perform any special work to generate a structured character string and a structured document using character input assist device 1. Furthermore, the use of a structured character string and a structured document facilitates connection to a database such as knowledge base 12.

[0080] Therefore, the structured character string and the structured document can be effectively and reliably used in various applications. Note that the kana character string "Kyou no Shibuya no Tenki wa Hare" and the structured character string or the structured document "Kyou no Shibuya no Tenki wa Sunny" shown here are merely examples, and the character input assistance device 1 can structure any input character string to generate a structured character string or a structured document. The basic functions and processing flow of the character input assistance device 1 have been described above.

[0081] [Variations] Next, a description will be given of a process for correcting a structured character string or a structured document generated by character input assistance device 1 as a first modified example of the present invention. Character input assistance unit 102 of character input assistance device 1 corrects a structured character string or a structured document in accordance with information indicating a user's operation input via input unit 100.

[0082] 7A to 7C are diagrams illustrating UI (user interface) screens used to correct a structured character string or a structured document. A user may correct a part or all of a structured character string or a structured document generated by character input support unit 102. In such a case, a user operation for correction is input to character input support unit 102 via input unit 100. In response to the input operation, for example, as shown in FIG. 7A, character input support unit 102 divides a structured character string or a structured document displayed as text "Today in Shibuya..." into words, and creates a character string "'Today' 'of' 'Shibuya' 'of'...".

[0083] The character input assist unit 102 outputs the character string thus delimited to the preview unit 104. The preview unit 104 displays the character string input from the character input assist unit 102 on a UI screen displayed on the display screen as shown in Fig. 7A, and presents it to the user.

[0084] Furthermore, when the user performs an operation on "Shibuya" from among the character strings included in the UI image shown in FIG. 7A, the character input assistance unit 102 causes this operation to be input via the input unit 100. The character input assistance unit 102 reads out the link "<#Shibuya (place name)>" included in the structured character string or structured document shown in FIG. 6. Furthermore, the character input assistance unit 102 outputs the word "Shibuya" that was the subject of the operation and its link "<#Shibuya (place name)>" to the preview unit 104. As shown in FIG. 7B, the preview unit 104 displays "Shibuya", which was the subject of the investigation, in the character string field on the UI screen, and displays the link "<#Shibuya (place name)>" in the link field.

[0085] When the user performs an operation to correct the word "Shibuya" to "Ikebukuro" and the link "#Shibuya (place name)" to "#Ikebukuro (station name)" on the UI screen shown in Fig. 7B as shown in Fig. 7C, these corrections are input to the character input support unit 102 via the input unit 100. As shown in Fig. 7C, the character input support unit 102 changes the character strings included in the structured character string or structured document, the character strings and links in the UI screen, from "Shibuya" to "Ikebukuro" and from "(place name) Shibuya" to "(station name) Ikebukuro" according to the user's correction operation.

[0086] Fig. 7D is a diagram illustrating the corrected structured character string or structured document. Fig. 7E is a diagram illustrating the entity of the knowledge base 12 corresponding to "(station name) Ikebukuro" included in the corrected structured character string or structured document. The character input support unit 102 corrects the link "<#Shibuya (place name)>" of "Shibuya" to the link "<#Ikebukuro (station name)>" of "Ikebukuro" as shown in bold with an underline in Fig. 7D among the structured character string or structured document shown in Fig. 6. Further, the character input support unit 102 outputs and displays the corrected structured character string or structured document on a display or the like. Also, the character input support unit 102 stores and saves the structured character string or structured document in the data storage unit 106.

[0087] Note that the entity including the link "<#Ikebukuro (station name)>" was included in the knowledge base 12 before the correction, and its content is as shown in Fig. 7E. That is, the entity including the link "<#Ikebukuro (station name)>" has the character strings "<schema:description" indicating Ikebukuro Station” and " <nextpart>It includes "<#Ikebukuro 1>" as structured attribute information. The entity containing the link "<#Ikebukuro 1>" has the character strings "<schema:name "池">", "<読み>"いけ"", "<schema:typicalAgeRange "7">", and " <nextpart>It includes "<#Ikebukuro 2>" as structured attribute information. The entity containing the link "<#Ikebukuro 2>" has the character strings "<schema:name "bag">", "<reading> "bukuro"", "<schema:typicalAgeRange "12">", and " <nextpart><#Ikebukuro (station name)>" is included as structured attribute information.

[0088] Next, a second modified example of the present invention will be described. In this modified example, the character input assist device 1 includes a structuring processing device 10 and a knowledge base 12 (FIG. 1), and further includes a subtitle display device (FIG. 8). The character input assist device 1 of this modified example can be used, for example, to generate subtitle data to be displayed on the screen of a broadcasting device (such as a television). FIG. 8 is a sequence diagram showing input and output of data for generating and displaying subtitles. The structuring processing device 10 (FIG. 1) generates a structured character string or a structured document, and stores and saves it in the data storage unit 106. The subtitle display device reads out a structured document file including a structured document from the data storage unit 106 of the character input assist device 1 (S140).

[0089] The subtitle display device processes the read structured document file and determines whether or not it is necessary to add readings (ruby) to the kanji (e.g., "Shibuya") in the structured document included in the structured document file (S142). The subtitle display device outputs an entity search request to the knowledge base 12 using the kanji (S144). The knowledge base 12 receives the search request from the subtitle display device, searches for an entity corresponding to the kanji used in the entity search request, and transmits the entity obtained as a search result to the subtitle display device. The subtitle display device receives the entity from the knowledge base 12 and obtains the readings of the kanji (e.g., "Shibuya") (S146).

[0090] The subtitle display device generates a reading (ruby) by associating the reading (e.g., "Shibuya") obtained by the process of S146 with the kanji (e.g., "Shibuya") used in the entity search request to the knowledge base 12. The subtitle display device generates text representing the structured character string or structured document received from the subtitle display device, and further generates subtitle data by adding a reading to the kanji included in the text (S148).

[0091] 9A and 9B are diagrams illustrating examples of subtitles generated by the subtitle display device. As shown in Fig. 9A, the subtitle display device transmits the generated subtitle data to the preview unit 104 via the character input assistance unit 102 of the character input assistance device 1, and as shown in Fig. 9A, the subtitle data in which pronunciations are added to all kanji characters is displayed as subtitles at the bottom of the display or the like (S150).

[0092] The subtitle display device converts the generated subtitle data into the ARIB-TTML, IMSC, or WebVTT format and outputs it to another device such as a broadcasting device (S152). Note that the subtitle data shown in Fig. 9A can be used by a broadcasting device to display subtitles on a television screen, for example.

[0093] In the entity shown in FIG. 2A, the link "<#Shibuya1>" contains the string "schema:typicalAgeRange"12"". By using this information in "schema:typicalAgeRange", it is possible to add readings to only some of the kanji characters included in the subtitles shown in FIG. 9A.

[0094] For example, a user of the subtitle display device can assume that the viewer who will read the subtitles is 8 years old or older and can cause the subtitle display device to generate subtitles with the reading kana "Shibuya" added only to the word "Shibuya" as shown in FIG. 9B. As shown in FIG. 2A and FIG. 2B, an entity with a link <#today> and corresponding to the word "today" includes a character string "schema:typicalAgeRange "8"" with a numerical value that sets that the age group requiring furigana is under 8 years old. Also, an entity with links <#weather1> and <#weather2> and corresponding to the word "weather" consisting of the kanji characters "heaven" and "ki" includes a character string "schema:typicalAgeRange "7"" with a numerical value that sets that the age group requiring furigana is under 7 years old. Also, an entity with a link <#shibuya1> and corresponding to the kanji character "shibu" in the word "Shibuya" includes a character string "schema:typicalAgeRange "12"" with a numerical value that sets that the age group requiring furigana is under 12 years old.

[0095] In this way, when a numerical value is assigned to the character string "schema:typicalAgeRange" included in each entity, the user sets the age of the viewer who will read the subtitles as 8 years or older in the subtitle display device. Then, in accordance with the numerical value of the character string "schema:typicalAgeRange" included in each entity and the age setting by the user, the subtitle display device generates and displays subtitles with pronunciation assigned only to the word "Shibuya" that contains the kanji character "shibu" whose numerical value in the entity's character string "schema:typicalAgeRange" is 12 (9 or higher).

[0096] The third modified example of the present invention will be described below. Fig. 10 is a diagram showing the configuration of the manuscript system screen generating device 3. As shown in Fig. 10, the manuscript system screen generating device 3 includes the structuring processing device 10 and knowledge base 12 shown in Fig. 1, a screen generating unit 30, and a weather forecast information source 32.

[0097] The knowledge base 12 includes entities related to traffic information before the start of the news broadcast. Furthermore, the knowledge base 12 generates entities related to weather forecast information from the weather forecast information source 32 before the start of the news broadcast. An input character string by a user creating a news manuscript is input via an input unit 100 to a character input support unit 102 of the structuring processing device 10 (FIG. 1), and the character input support unit 102 generates a structured character string or a structured document of the news manuscript. The generated structured character string or structured document is output to a screen generation unit 30, which receives the structured character string or structured document.

[0098] In the manuscript system screen generating device 3, the screen generating unit 30 outputs a search request for entities to the knowledge base 12 using a plurality of kana-kanji character strings included in the structured character string or structured document of the news manuscript generated by the character input support unit 102. The knowledge base 12 receives the search request from the screen generating unit 30, and in response to this search request, transmits to the screen generating unit 30 entities of traffic information and weather forecast information that correspond to the "place names" and "station names" that are included in the structured character string or structured document of the news manuscript and that are associated with each other as described above.

[0099] The screen generating unit 30 generates news information showing, as text, the structured character string or structured document of the news manuscript received from the structuring processing device 10, and traffic information and weather forecast information related to a place name, for example, "Ikebukuro," received from the knowledge base 12. Furthermore, the screen generating unit 30 generates a manuscript system screen including the generated news information, traffic information, and weather forecast information. In other words, according to the manuscript system screen generating device 3, not only information related to "Ikebukuro" as a place name included in the news manuscript, but also information related to "Ikebukuro" as a station name can be provided to, for example, an announcer appearing on a news program and reading out the news manuscript.

[0100] Fig. 11 is a diagram illustrating an example of a manuscript system screen generated by the screen generation unit 30 of the manuscript system screen generation device 3 shown in Fig. 10. Fig. 11 shows a case where "Ikebukuro (place name)" and "Ikebukuro (station name)" are associated with each other by an entity included in the knowledge base 12. As shown in Fig. 11, a screen showing, as text, a structured character string or structured document of news of an event being held in "Ikebukuro" as a place name is displayed in the left half of the manuscript system screen. The manuscript system screen is presented to the announcer as described above.

[0101] Meanwhile, a screen is displayed in the upper right corner of the manuscript system screen showing information about delays occurring on a specific railway line that passes through the station name "Ikebukuro" and that railway lines other than the specific line are operating normally. Furthermore, a screen is displayed in the lower right corner of the manuscript system screen showing weather forecast information in the vicinity of the place name "Ikebukuro". Note that the characters displayed on the manuscript system screen may be color-coded as necessary, for example, only information about railway lines that are experiencing some kind of abnormality among the railway lines that pass through the station name "Ikebukuro" may be displayed in red.

[0102] By viewing the screen of the script system, the announcer appearing on the news program can notice if there is any important information in the traffic or weather forecast related to the place name included in the news script, and therefore the announcer can read out the important information that occurs in the traffic or weather forecast as appropriate.

[0103] The processing by the character input assist device according to the present invention can be realized by software as described above. When this processing is realized by software, the program constituting this software is installed and executed on the user's computer. Such a program can be distributed and installed on the user's computer by removable media, or distributed and installed on the user's computer from another computer via a network. Furthermore, the program constituting this software does not necessarily have to be distributed to the user's computer, and can be provided to the user's computer as a Web service, for example.

[0104] Although the present invention has been described in detail, the present invention is not limited to the above-mentioned individual embodiments. Various additions, replacements, modifications, partial deletions, etc. may be made to these embodiments without departing from the gist of the present invention, or without departing from the gist of the present invention derived from the contents described in the claims and their equivalents. These embodiments may also be implemented in combination. For example, in the above-mentioned embodiments, the order of each operation and the order of each process are shown as examples, and are not limited to these. The same applies when numerical values ​​or formulas are used in the description of the above-mentioned embodiments. [Explanation of symbols]

[0105] 1. Character input support device 10 Structure processing device (structure processing unit) 100 Input section 102 Text input support section 104 Preview section 106 Data Storage Section 12 Knowledge Base 3. Manuscript system screen generator 30 Screen generation section 32 Weather forecast sources< / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart> < / nextpart>

Claims

1. an input section into which an input string is input; a knowledge base that returns, in response to a search request using the input string, one or more entities including attribute information corresponding to the input string; a character input support unit that receives one or more entities returned by the knowledge base, outputs the entities, structures the input character string in response to a user's operation of selecting one of the output entities, and outputs the structured input character string; a preview unit that displays the entity output by the character input support unit; a data storage unit for storing the structured input character string; A character input assistance device comprising:

2. the attribute information includes information corresponding to the input character string; the knowledge base returns the entities containing information corresponding to the input string used in the search request; The input support device according to claim 1 .

3. the character input support unit sets the input entity as a conversion candidate for the input character string and outputs the entity to the preview unit; The character input assist device according to claim 1 .

4. A user's operation of selecting one of the output entities is an operation of identifying one of the conversion candidates of the input character string, the preview unit receives conversion candidates for the input character string from the character input support unit and displays them; The character input support unit receives an operation for identifying one of conversion candidates of the input character string via the input unit. The character input assist device according to claim 3 .

5. the character input assistance unit structures the input string by adding the entity returned by the knowledge base to the structured input string in response to an operation of identifying one of conversion candidates for the input string; The character input assist device according to claim 4.

6. a knowledge base that returns an entity including attribute information corresponding to an input string in response to a search request using the input string; A structuring processor; A character input assistance system comprising: The structuring processing device includes: an input unit for inputting the input character string; a character input support unit that receives one or more entities returned by the knowledge base in response to the search request using the input character string, outputs the entities, structures the input character string in response to a user operation of selecting one of the output entities, and outputs the structured input character string; a preview unit that displays the entity output by the character input support unit; a data storage unit for storing the structured input character string; Equipped with Character input support system.

7. A step of inputting an input string; In response to a search request using the input string, returning one or more entities including attribute information corresponding to the input string; outputting one or more entities returned in response to the search request, structuring the input string input in response to a user operation of selecting one of the output entities, and outputting the structured input string; displaying the outputted entities; storing the structured input string; A program that causes a computer to execute the following.

Images