JP2010271870A - Analysis device for consecutive pictographs or the like - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To create a dictionary based on actual usage for consecutive pictographs or the like. <P>SOLUTION: An analysis device for consecutive pictographs or the like includes: a means for extracting consecutive-pictographs or the like, sentence and article that refers to a single-pictographs-or-the-like dictionary to extract, from obtained content data, consecutive pictographs or the like, a sentence including the consecutive pictographs or the like, and an article including the sentence; an article feature word extraction means that refers to a word dictionary to extract an article feature word from the extracted article; a means for extracting a sentence excluding consecutive pictographs or the like that extracts another sentence matching a portion excluding the consecutive pictographs or the like of the extracted sentence from the obtained content data; a sentence feature word extraction means that refers to the word dictionary to extract a sentence feature word from the another extracted sentence; a feature word similarity determination means that refers to a similar word dictionary to determine whether or not the extracted article feature word is similar to the sentence feature word; and a means for registering consecutive pictographs or the like data that, if they are determined to be similar, associates the extracted consecutive ictographs or the like with the sentence feature word to register them in a consecutive-pictographs-or-the-like dictionary. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a technique for analyzing pictograms used for blogs (diary-like sites) on the Internet, bulletin boards, and the like.

  Pictographs based on illustration characters and emoticons expressing human faces by arranging normal characters are often used in e-mails, blogs, bulletin boards, and the like.

  Many of the character conversion functions installed in mobile phones and the like support single pictograms and emoticons, and it is possible to easily input pictograms and emoticons during the text input process.

  Patent Document 1 discloses a technique for analyzing the type of a pictograph (emoticon) and reproducing an animation image or sound corresponding to the pictograph in synchronization with the display of text data.

JP 2008-54340 A

  As described above, for each pictogram or emoticon, a conversion dictionary has been developed. However, recently, a specific meaning (word) has been expressed by continuing a plurality of pictograms and emoticons. Here, such a series of a plurality of pictograms is called “continuous pictograms”. A single pictogram or emoticon is referred to as “pictogram etc.”.

  For such continuous pictographs, the conversion dictionary has not been maintained, and the user has to input characters one by one. Further, when analyzing content on the server side, meanings such as continuous pictographs cannot be interpreted, and therefore effective content analysis cannot be performed.

  The present invention has been proposed in view of the above-described conventional problems, and an object of the present invention is to provide a continuous pictograph analysis device capable of generating a dictionary for continuous pictographs based on how they are actually used. Is to provide.

  In order to solve the above-described problems, in the present invention, as described in claim 1, content acquisition means for acquiring content data from a content database to be analyzed, and the acquired content data alone With reference to the pictogram dictionary, continuous pictograms, texts containing the continuous pictograms, etc., continuous pictograms, text / article extraction means for extracting articles containing the text, and word dictionary from the extracted articles Article feature word extraction means for extracting article feature words with reference to the above, and consecutive pictograms excluded text that extracts other text that matches the part of the extracted text excluding continuous pictograms etc. from the acquired content data Extraction means, sentence feature word extraction means for referring to a word dictionary from other extracted sentences to extract sentence feature words, and extracted article feature words and sentence feature words A feature word similarity determination unit that determines whether or not they are similar by referring to the synonym dictionary, and if it is determined to be similar, the extracted continuous pictograms and the text feature words are associated with each other and registered in the continuous pictogram dictionary. The gist is a continuous pictograph etc. analyzing device comprising continuous pictograph data registering means.

  In addition, as described in claim 2, in the continuous pictograph etc. analyzing device according to claim 1, the number of appearances is updated based on the record of the number of appearances of the continuous pictographs with reference to the dictionary of continuous pictograms, etc. When the image is not performed for a certain period of time, it is possible to provide continuous pictogram data deletion means for deleting the continuous pictogram or the like.

  In addition, as described in claim 3, a content acquisition step of acquiring content data from a content database to be analyzed, a continuous pictogram, etc. with reference to a single pictogram dictionary from the acquired content data, Article feature words for extracting article feature words by referring to a word dictionary from the extracted articles and continuous pictograms, sentences, and article extraction process for extracting sentences including the continuous pictograms and articles including the sentences An extraction step, an extracted sentence extraction step for extracting other sentences that match a portion of the extracted sentence excluding continuous pictograms, etc., and a word dictionary from the extracted other sentences Text feature word extraction process that extracts text feature words by referring to the text, and whether or not they are similar by referring to the synonym dictionary from the extracted article feature words and text feature words A continuous pictogram or the like including a feature word similarity determination step and a continuous pictogram or the like data registration step for registering the extracted continuous pictogram or the like and the text feature word in association with the dictionary in the continuous pictogram dictionary when determined to be similar It can be configured as an analysis method.

  In the continuous pictogram analyzing apparatus of the present invention, a dictionary can be generated based on the actual usage of continuous pictograms and the like, and can be used for input and analysis of continuous pictograms and the like.

It is a figure which shows the structural example of the system concerning one Embodiment of this invention. It is a figure which shows the data structure example of a content database. It is a figure which shows the data structure example of a dictionary, such as a single pictogram. It is a figure which shows the data structure example of a word dictionary. It is a figure which shows the example of a data structure of a synonym dictionary. It is a figure which shows the example of a data structure of the database for extraction work. It is a figure which shows the example of a data structure of a dictionary, such as a continuous pictograph. It is a figure which shows the hardware structural example of analysis servers, such as a pictograph. It is a sequence diagram (the 1) which shows the process example of embodiment. It is a sequence diagram (the 2) which shows the process example of embodiment. It is a figure which shows the example of text / article extraction containing continuous pictograms, etc. and article feature word extraction. It is a figure which shows the example of exclusion sentence extraction and pictorial feature word extraction, such as a pictograph.

  Hereinafter, preferred embodiments of the present invention will be described.

<Configuration>
FIG. 1 is a diagram showing a configuration example of a system according to an embodiment of the present invention.

  In FIG. 1, a plurality of user terminals 2 such as a PC (Personal Computer), a mobile phone, and a PDA (Personal Digital Assistants) operated by a user are connected to a network 1 such as the Internet. The user terminal 2 includes a general browser (Web browser) 21. The browser 21 performs request / acquisition / display of page data described in a language such as HTML (Hyper Text Markup Language), transmission of form data, and the like according to HTTP (Hyper Text Transfer Protocol) which is a standard protocol of the Internet. It has a function.

  On the other hand, connected to the network 1 is a blog / bulletin board server 3 for managing browsing / posting of blogs / bulletin boards in response to access from the browser 21 of the user terminal 2 operated by the user. The blog / bulletin board server 3 is provided with a content database 301. This database systematically holds predetermined data on a storage medium such as an HDD (Hard Disk Drive) in a computer that holds the database.

  The network 1 is connected to a pictograph analysis server 4 that analyzes the contents of articles on the blog / bulletin board server 3.

  The pictographic etc. analysis server 4 includes a blog / bulletin board content acquisition unit 41, a continuous pictograph etc./text/article extraction unit 42, an article feature word extraction / region determination unit 43, a continuous pictograph etc. excluded text extraction unit 44 and a text as functional units. A feature word extraction unit 45, a feature word similarity determination unit 46, a continuous pictograph data registration unit 47, and a continuous pictograph data deletion unit 48 are provided. These functional units are realized by a computer program executed on hardware resources such as a CPU (Central Processing Unit), a ROM (Read Only Memory), and a RAM (Random Access Memory) of the computer constituting the pictogram analysis server 4. It is what is done. These functional units do not need to be arranged on a single computer, and may be distributed as necessary.

  Further, as a dictionary or database that the pictogram analysis server 4 refers to or updates, a single pictogram dictionary 401, a word dictionary 402, a synonym dictionary 403, an extraction work database 404, and a continuous pictogram dictionary 405 are provided. These dictionaries and the like systematically hold predetermined data on a storage medium such as an HDD in a computer that holds the dictionaries and the like.

  FIG. 2 shows an example of the data structure of the content database 301 of the blog / bulletin board server 3. The content database 301 includes items such as “registration date”, “registrant user ID”, “registrant IP address”, and “article content”. “Registration date” is the date and time when the article was registered (published). The “registrant user ID” is the user ID of the user who registered the article. The “registrant IP address” is the IP address of the transmission source at the time of registration of the user who registered the article. “Article content” is data of the content of an article, and includes text, images, sounds, and the like.

  FIG. 3 is a diagram showing an example of the data structure of the single pictogram dictionary 401 of the pictogram analysis server 4. The single pictogram dictionary 401 includes items such as “pictogram etc.”, “code”, and “word”. “Emoji” or the like is an emoticon representing one character or one meaning. “Code” is information for identifying a pictograph or the like. “Word” is a word corresponding to a pictograph or the like.

  FIG. 4 is a diagram showing an example of the data structure of the word dictionary 402, and has items such as “word” and “part of speech”. A “word” is a word that becomes a headline. “Part of speech” is the part of speech of the word.

  FIG. 5 is a diagram showing an example of the data structure of the synonym dictionary 403, and has items such as “word” and “synonym”. A “word” is a word that becomes a headline. A “synonym” is other word (s) that are similar to the word.

  FIG. 6 is a diagram showing an example of the data structure of the extraction work database 404, and has items such as “continuous pictograms,” “words,” “appearance count (update date and time)”, “region”, and the like. “Continuous pictograms” are a plurality of continuous pictograms or emoticons. A “word” is a word corresponding to a continuous pictograph or the like. “Number of appearances (update date / time)” is the number of times of appearance by the analysis and the last update date / time. The “region” is a region where a user who has written an article in which a continuous pictograph or the like appears is located.

  FIG. 7 is a diagram showing an example of the data structure of the continuous pictogram dictionary 405, and includes items such as “continuous pictograms”, “words”, “appearance count (update date and time)”, “region”, and the like. The items are the same as those in the extraction work database 404 of FIG.

  Returning to FIG. 1, the blog / bulletin board content acquisition unit 41 of the pictogram etc. analysis server 4 has a function of acquiring content data from the content database 301 of the blog / bulletin board server 3 during the pictogram etc. analysis processing.

  The continuous pictogram / text / article extraction unit 42 refers to the single pictogram dictionary 401 from the content data acquired by the content database 301, the text including the continuous pictogram, the continuous pictogram, etc., and the text It has a function to extract articles including.

  The article feature word extraction / region determination unit 43 has a function of referring to the word dictionary 402 from the articles extracted by the continuous pictograms, sentences / article extraction unit 42 and extracting the feature words. The article feature word extraction / area determination unit 43 also has a function of determining the area where the registrant is located based on text included in the article, registrant IP address, registrant user ID, and the like.

  The consecutive pictogram etc. excluded text extraction unit 44 matches the content data acquired by the content database 301 based on the part excluding the continuous pictogram etc. from the text extracted by the continuous pictogram etc./text/article extraction unit 42. It has a function to extract sentences.

  The sentence feature word extraction unit 45 has a function of referring to the word dictionary 402 from the sentence extracted by the consecutive pictogram etc. exclusion sentence extraction unit 44 and extracting a feature word.

  The feature word similarity determination unit 46 uses the feature words (article feature words) extracted by the article feature word extraction / region determination unit 43 and the feature words (sentence feature words) extracted by the sentence feature word extraction unit 45 as synonyms. It has a function of referring to the dictionary 403 to determine whether or not they are similar. Also, if the feature word similarity determination unit 46 determines that they are similar, the feature word similarity determination unit 46 temporarily registers continuous pictograms or the like in the extraction work database 404, and if the number of appearances exceeds a predetermined value, the main word registration data registration unit 47 It also has a function to instruct

  The continuous pictogram data registration unit 47 has a function of registering a continuous pictogram or the like in the continuous pictogram dictionary 405 when receiving an instruction for main registration from the feature word similarity determination unit 46.

  The continuous pictogram data deletion unit 48 has a function of referring to the extraction work database 404 and the continuous pictogram dictionary 405 at an appropriate timing and deleting the continuous pictograms determined to be abandoned.

  FIG. 8 is a diagram illustrating a hardware configuration example of the pictogram analysis server 4.

  In FIG. 8, the pictogram analysis server 4 is connected to a CPU 4002, ROM 4003, RAM 4004, NVRAM (Non-Volatile Random Access Memory) 4005, I / F (Interface) 4006, and I / F 4006 connected to a system bus 4001. Also provided are an input / output device (I / O) 4007 such as a keyboard, mouse, monitor, CD / DVD (Compact Disk / Digital Versatile Disk) drive, HDD 4008, NIC (Network Interface Card) 4009, and the like. M is a medium (recording medium) such as a CD / DVD in which a program or data is stored.

<Operation>
FIG. 9 and FIG. 10 are sequence diagrams showing a processing example of the above embodiment.

  In FIG. 9, when the user of the user terminal 2 makes a posting on a blog or bulletin board, the blog / bulletin board server 3 is accessed from the browser 21 of the user terminal 2 to make a posting page request (step S101). The page request is made by transmitting a message including a GET method according to HTTP, a standard protocol of the Internet, a request URI (Uniform Resource Indicator), and the like from the browser 21 of the user terminal 2 to the blog / bulletin board server 3.

  In response to this, the blog / bulletin board server 3 returns the page data of the posting page held inside or dynamically generated to the browser 21 of the user terminal 2 (step S102). The page data is described in HTML or the like, and is transmitted from the blog / bulletin board server 3 to the browser 21 of the user terminal 2 in accordance with an HTTP response or the like.

  The browser 21 of the user terminal 2 displays the posting page (step S103), and the user inputs the posting content (step S104).

  When the input to the posting page is completed, the posting content is transmitted from the browser 21 of the user terminal 2 to the blog / bulletin board server 3 (step S105). The input content is transmitted from the browser 21 of the user terminal 2 to the blog / bulletin board server 3 as data such as HTML added to the HTTP POST method or PUT method, or as a parameter added to the GET method.

  The blog / bulletin board server 3 that has received the posted content registers the content of the article (posted content) in the content database 301 (FIG. 2) in association with the registration date, registrant user ID, registrant IP address, and the like (step S106). .

  Thereafter, at an appropriate timing, the pictograph analysis server 4 starts its operation, and the blog / bulletin board content acquisition unit 41 acquires content data from the content database 301 of the blog / bulletin board server 3 (step S111). The content data to be acquired can be limited to data having a predetermined freshness in consideration of the registration date and time since old data is not helpful. The acquired content data is temporarily held in the inside so that the continuous pictograph / text / article extracting unit 42 and the continuous pictograph excluded text extracting unit 44 can be used (steps S112 and S113).

  Next, the continuous pictogram / text / article extraction unit 42 refers to the single pictogram dictionary 401 from the content data acquired by the content database 301, and includes continuous pictograms and the text including the continuous pictograms, An article including the sentence is extracted (step S114). That is, whether the entire content data is searched using the pictograms registered in the single pictogram dictionary 401 as a key, or whether the pictograms registered in the single pictogram dictionary 401 appear by scanning the content data Judgment is made, and the continuous pictograms and the like that have appeared, the text including the continuous pictograms, and the article including the text are extracted. Note that the continuous pictograms extracted at the time of the previous analysis can be prevented from being duplicated in consideration of the registration date and time of the content data. FIG. 11 shows a state in which a sentence S1 including continuous pictograms L1 and an article A1 including the sentence S1 are extracted.

  Returning to FIG. 9, the continuous pictograph / text / article extraction unit 42 delivers the extracted continuous pictograph, etc. (code), text, and article to the article feature word extraction / region determination unit 43 (step S <b> 115).

  Next, the article feature word extraction / region determination unit 43 refers to the word dictionary 402 from the articles extracted by the continuous pictograms / text / article extraction unit 42 and extracts feature words (article feature words) (step). S116). In the extraction of feature words, for example, a character string included in an article and a word registered in the word dictionary 402 are compared and determined by the longest match method or the like, and a word with a high matching frequency is determined as a feature word. . In the case of the article A1 shown in FIG. 11, for example, “first grader” is extracted as the feature word C1.

  Returning to FIG. 9, the article feature word extraction / region determination unit 43 determines the region where the registrant is located based on the text included in the article, the registrant IP address, the registrant user ID, and the like (step S117). That is, when a word indicating a place name is included in the text included in the article, it is set as a region. If the registrant IP address is included in advance in the IP address range of a table (not shown) in which the IP address range is associated with the region, the corresponding region is acquired. If the user DB (not shown) is referred to from the registrant user ID and the address can be acquired from the profile, the address is set as the area.

  Next, the article feature word extraction / region determination unit 43 converts the continuous pictograms, etc., the continuous pictograms, etc. passed from the sentence / article extraction unit 42, and the regions determined as the article feature words extracted by itself into continuous pictograms, etc. Delivered to the excluded text extraction unit 44 (step S118).

  Next, the consecutive pictograms excluded text extraction unit 44, from the content data acquired by the content database 301, based on the part excluding the continuous pictograms etc. from the text extracted by the continuous pictograms, text / article extraction unit 42, A matching sentence is extracted (step S119). That is, a character string excluding continuous pictograms and the like is generated from the character string of the text extracted by the continuous pictogram / text / article extraction unit 42, and the entire content data is searched using the character string as a key, Alternatively, the content data is scanned to determine whether or not the character string appears, and a sentence including the character string is extracted. A character string “Congratulations” is generated from the sentence S1 of FIG. FIG. 12 shows a state in which sentences S2 and S3 are extracted from the content data based on the character string “Congratulations”.

  Returning to FIG. 9, the consecutive pictograms and other excluded text extraction unit 44 extracts the continuous pictograms, the article feature words, and the region passed from the article feature word extraction / region determination unit 43, and the sentence feature word extraction. Delivered to the unit 45 (step S120).

  Next, in FIG. 10, the sentence feature word extraction unit 45 refers to the word dictionary 402 from the sentence extracted by the consecutive pictogram etc. exclusion sentence extraction unit 44 and extracts a feature word (sentence feature word) (step S <b> 121). ). In the extraction of feature words, for example, a character string included in a sentence and a word registered in the word dictionary 402 are compared and determined by a longest match method or the like, and a word with a high matching frequency is determined as a feature word. . In the case of the sentence S2 shown in FIG. 12, the character string “entrance ceremony” included in the sentence S2 is extracted as the feature word C2. In the case of the sentence S3, “birthday” included in the sentence S3 is extracted as the feature word C3.

  Returning to FIG. 10, the sentence feature word extraction unit 45 converts the continuous pictograms, the article feature words, the region, and the sentence feature words extracted from the continuous pictogram etc. excluded sentence extraction unit 44 into the feature word similarity determination unit. 46 (step S122).

  Next, the feature word similarity determination unit 46 refers to the synonym dictionary 403 for the article feature words extracted by the article feature word extraction / region determination unit 43 and the sentence feature words extracted by the sentence feature word extraction unit 45. It is determined whether or not they are similar (step S123). That is, the synonym dictionary 403 is referenced using one feature word as a key, and it is determined whether or not the other feature word is similar depending on whether it is registered as a synonym. Between the feature word C1 “first grader” extracted in FIG. 11 and the feature word C2 “entrance ceremony” and feature word C3 “birthday” extracted in FIG. 12, “first grader” and “entrance ceremony” Are similar.

  Returning to FIG. 10, when the feature word similarity determination unit 46 determines that they are similar, the feature word similarity determination unit 46 temporarily registers continuous pictograms and the like in the extraction work database 404 for the combinations determined to be similar (step S 124). That is, if the sentence feature word is a word and the extraction work database 404 already has the same continuous pictograph and the like and a record of the word and the region, the appearance count is incremented by one. If there is no record, a record of continuous pictograms, words, and regions is created, and the number of applications is set to 1.

  Next, the feature word similarity determination unit 46 determines whether or not the number of appearances exceeds a predetermined value (step S125). If the number of appearances exceeds the predetermined value, the continuous pictograms, the word, the number of appearances, and the region are determined. The data is handed over to the continuous pictograph data registration unit 47 to instruct main registration (step S126).

  In response, the continuous pictogram data registration unit 47 registers the continuous pictogram etc., the word, the number of appearances, and the region in the continuous pictogram dictionary 405 (step S127). If there is already a record with the same word and area as the continuous pictograph etc. other than the appearance count, only the appearance count is updated.

  On the other hand, at an appropriate timing thereafter, the continuous pictogram data deletion unit 48 refers to the extraction work database 404 and the continuous pictogram dictionary 405 (step S131), and determines the continuous pictograms to be deleted that are determined to be obsolete. (Step S132). That is, the number of appearances does not increase over a predetermined period because the update date and time has passed for a predetermined period or more, and other records having the same continuous pictograph or the like exist as deletion targets. In this case, the same continuous pictogram record means the same including the region. This is because it is normal that different pictographs used in different meanings coexist in different regions.

  Then, the continuous pictograph data registration unit 47 deletes the corresponding record from the extraction work database 404 and the continuous pictograph dictionary 405 (step S133).

<Summary>
As described above, according to the present embodiment, the continuous pictogram dictionary 405 can be kept up-to-date based on the actual usage of continuous pictograms and the like in the blog / bulletin board server 3. It can be used effectively for input and analysis.

  The present invention has been described above by the preferred embodiments of the present invention. While the invention has been described with reference to specific embodiments, various modifications and changes may be made to the embodiments without departing from the broad spirit and scope of the invention as defined in the claims. Obviously you can. In other words, the present invention should not be construed as being limited by the details of the specific examples and the accompanying drawings.

DESCRIPTION OF SYMBOLS 1 Network 2 User terminal 21 Browser 3 Blog / bulletin board server 301 Content database 4 Pictogram etc. analysis server 41 Blog / bulletin board content acquisition part 42 Continuous pictogram etc./Sentence/article extraction part 43 Article feature word extraction / region determination part 44 Continuous pictogram etc. Excluded sentence extraction section 45 Sentence feature word extraction section 46 Feature word similarity determination section 47 Continuous pictogram data registration section 48 Continuous pictogram data deletion section 401 Single pictogram dictionary 402 Word dictionary 403 Synonym dictionary 404 Extraction work database 405 Continuous pictogram, etc. dictionary

Claims (3)

Content acquisition means for acquiring content data from a content database to be analyzed;
With reference to a dictionary of single pictograms, etc., from the acquired content data, continuous pictograms, etc., continuous pictograms etc., text / article extraction means for extracting sentences containing the continuous pictograms, etc., and articles containing the text, and ,
Article feature word extraction means for extracting an article feature word from the extracted article by referring to a word dictionary;
Excluded sentence extraction means such as continuous pictograms for extracting other sentences that match the part of the extracted text excluding the continuous pictograms of the extracted text,
A sentence feature word extracting means for extracting a sentence feature word from another extracted sentence by referring to a word dictionary;
Feature word similarity determination means for determining whether or not they are similar by referring to a synonym dictionary from the extracted article feature words and sentence feature words;
A continuous pictogram etc. analyzing device characterized by comprising continuous pictogram etc. data registration means for registering the extracted continuous pictogram etc. and the sentence feature word in association with each other when it is determined to be similar . In the analysis apparatus for continuous pictograms according to claim 1,
With reference to the continuous pictogram dictionary, based on the record of the number of appearances of continuous pictograms, and the like, provided with a continuous pictogram data deletion means for deleting the continuous pictograms and the like when the number of appearances is not updated for a certain period of time. A device for analyzing continuous pictograms, etc. A content acquisition process for acquiring content data from a content database to be analyzed;
From the acquired content data, referring to a dictionary of single pictograms, continuous pictograms, etc., continuous pictograms etc., text / article extraction process for extracting texts containing the continuous pictograms, etc., and articles containing the texts, ,
An article feature word extraction step of extracting an article feature word from the extracted article by referring to a word dictionary;
From the acquired content data, an extracted sentence extraction process such as continuous pictograms that extracts other sentences that match the part of the extracted text that excludes continuous pictograms, and the like,
A sentence feature word extraction step of extracting a sentence feature word from another extracted sentence with reference to a word dictionary;
A feature word similarity determination step for determining whether or not they are similar by referring to a synonym dictionary from the extracted article feature words and sentence feature words;
A continuous pictogram etc. data registration method comprising: a continuous pictogram etc. data registration step of registering the extracted continuous pictogram etc. and a text feature word in association with each other in a dictionary of continuous pictograms, etc., when determined to be similar .

Images