JP2011100191A - Device, method, and program for retrieving document - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a user with a retrieval result of suitable ranking according to a user's decoding level. <P>SOLUTION: When a query (a retrieval word) is input from a user terminal 12, a retrieval engine part 16 retrieves an electronic document group based on the query, and sets basic ranking to a plurality of electronic documents included in the retrieval according to a degree of association with the query. A difficulty degree giving part 17 gives a degree of difficulty associated with the decoding level, to each of the plurality of electronic documents included in the retrieval using a difficulty degree dictionary 18. A ranking adjusting part 23 adjusts ranking according to the decoding level and the degree of difficulty so that the electronic document with the degree of difficulty corresponding to the user's decoding level preferentially has the high ranking. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an electronic document search technique, and more particularly to a technique for assigning an appropriate ranking to a plurality of electronic documents included in a search result according to a user's decoding level.

  As is well known, a query (also called a search term or keyword) is input to a server computer connected via the Internet using a client computer such as a mobile terminal or a personal computer, and a search result related to the query from an electronic document group. The document retrieval system for acquiring the URL is currently used by many users.

  In such a document search system, in order to obtain information that the user is aiming for, the user specifies a query that seems to be related to the information, and in general, the user wants the information. Search processing is performed regardless of the user's experience, knowledge and language level, etc., and multiple electronic documents related to the query are returned and displayed to the user in a predetermined ranking (display order) . For this reason, for example, when the difficulty of the electronic document displayed is high compared to the decoding level of the user, such as when the user is a child, even if the electronic document is highly relevant to the search term, There arises a problem that the contents of the electronic document cannot be understood or it takes time to deepen the understanding.

  On the other hand, it is also known that, due to security problems, unfavorable types of sites are registered in categories and returned with a filter such as not presenting to children. For example, Patent Document 1 discloses a technique related to processing. In Patent Document 2, interactive information exchange is performed with a user, a search result is inspected based on the information of the user, inappropriate information is filtered out, and a search result is sent to the user. Techniques for providing are described.

Japanese Patent No. 3605343 JP 2001-222538 A

  However, as described above, in the case where the search process is performed regardless of the user's decoding level, if the user's decoding level is lower than the difficulty level of the document, the document contains useful information. Even if, for example, the user is a child, the content of the document is too difficult to understand, or specialized terms and new phrases are used frequently, so it is time to deepen understanding for non-specialist users. There was a problem such as requiring.

  In addition, as a result of a search by a user who does not belong to a given community, if a document that uses a lot of words and phrases that can be understood only within a certain community is obtained as a result, it may be considered that the content cannot be understood. It is done. For example, when a user with other attributes reads high school student community information, or when a user with little knowledge of a computer is familiar with the computer, Even if a document is given as a result, it may not be understood without understanding the word.

  On the other hand, in the case where the filtering process is performed as described in Patent Document 1 and Patent Document 2, for example, when a game site is registered as a site category that is not preferable for a child, as a result of the child searching for game software, Even if software information is included, if the site category is a game site, it is not included in the search results, and there is a possibility that useful information cannot be provided to the user.

  The present invention has been made in view of such a problem, and an object of the present invention is to provide search results to the user with an appropriate ranking (display order) according to the decoding level of the user.

  Therefore, in one aspect of the present invention, search means for searching for an electronic document group based on a query input by a user and setting a ranking for the electronic document included in the search result according to the degree of relevance to the query; A decryption level determination means for determining a decryption level of the user, a difficulty level giving means for assigning a difficulty level related to the decryption level to at least an electronic document included in the search result, and the difficulty level and the decryption level And ranking adjustment means for preferentially positioning an electronic document having a difficulty level corresponding to the level of decoding according to the ranking.

  The “decoding level” is set as a numerical value (0 to 100%) or a stepped level (child, student, adult, etc.) according to individual parameters such as age, occupation, gender, etc. Alternatively, it can be estimated based on a search result, a browsing time, or the like. “Difficulty” corresponds to the user's decoding level. Like the decoding level, the “difficulty level” depends on individual parameters such as age, occupation, gender, etc. Student, adult, etc.).

  As described above, in the present invention, among the plurality of electronic documents searched according to the query, the electronic documents having the difficulty level according to the level of decoding such as the user's experience, knowledge and language level are preferentially ranked high. It is possible to return and present it to the user in this way, and the efficiency for the user to decode and understand the information can be greatly improved.

  In addition, since the ranking is adjusted according to the user's decoding level, some electronic documents are not excluded from the search results as in the filter processing described above. For this reason, even if the electronic document with the difficulty level that matches the decryption level is preferentially placed at the top of the ranking, the ranking of the electronic document containing useful information may be lowered, but the user can By sequentially referring to the document, an electronic document including this useful information can be found relatively easily.

  Preferably, the difficulty level assigning unit further searches the electronic document group to acquire information related to the difficulty level, and generates a difficulty level dictionary based on the information related to the difficulty level. The difficulty level dictionary generating means for performing the difficulty level determination means for determining the difficulty level of each electronic document using the difficulty level dictionary. Thus, by using the difficulty level dictionary, it is possible to determine the difficulty level of each electronic document more accurately and quickly.

  The present invention may be in the form of a program that causes a computer to function as the document search apparatus described above, and this program may be recorded in a recording medium.

  According to the present invention, the user can adjust the ranking of the plurality of electronic documents included in the search result according to the user's decoding level and the difficulty level of each of the plurality of electronic documents included in the search result. It is possible to preferentially provide an electronic document having a difficulty level corresponding to the level of decryption. As a result, the efficiency with which the user decodes and understands the information can be greatly improved.

1 is a configuration diagram showing a document search apparatus according to an embodiment of the present invention. The block diagram which shows the processing content of the difficulty level provision part of the said embodiment. The block diagram which shows the processing content of the search engine part and ranking adjustment part of the said embodiment.

  A document search apparatus according to an embodiment of the present invention will be described below with reference to the drawings.

  Referring to FIG. 1, this document search apparatus 10 constitutes a system of a Web search engine (so-called search engine) connected to a user terminal 12 through a network such as the Internet so as to be able to send and receive data.

  The user terminal 12 is configured by a mobile terminal such as a personal computer (PC) 12A or a mobile phone 12B, for example. A browser that displays a search interface unit 24 for document search and a query (search keyword group) input to the search interface 24 What is necessary is just to have input parts, such as a keyboard and a touch panel which perform.

  Specifically, the document search apparatus 10 includes hardware resources of a normal computer such as a CPU, a memory (RAM), a hard disk drive device, a communication device, and the like. As a result of cooperation between hardware resources such as one or a plurality of servers and software resources, the document search apparatus 10 performs each of the units 13 to 19 for performing electronic document group search processing, search result ranking setting / adjustment processing, and the like. 23 and 24 are implemented. In FIG. 1, as an example, a pre-processing 11 </ b> A for obtaining the degree of difficulty in which each unit 13 to 19, 23, 24 is used to adjust search engine index information and search result ranking from a search target electronic document group, and a user This is broadly divided into search request input processing 11B that adjusts the ranking of search results when a search processing command is issued from the terminal 12.

  More specifically, the document search apparatus 10 provides a document collection unit 13 that collects documents from the Internet and index information such as phrases and frequencies included in the collected documents to each document as components of the search engine. A search data storage unit 15 that stores index information such as words and phrases included in each document and a search request (search request by a user) ) To search for multiple electronic documents related to the search term (query), and set the basic ranking corresponding to the degree of matching / relevance with the search term for the multiple electronic documents included in the search results And a search engine unit 16. The functions and configurations of such a search engine are well known, and detailed description is omitted here, but an existing search engine such as Google (registered trademark) can be used simply.

  Further, the document search device 10 generates / updates the difficulty level dictionary 18, and uses the difficulty level dictionary 18 to give a difficulty level assigning unit that gives a difficulty level corresponding to the decryption level to each electronic document. 17.

  As shown in FIG. 2, the difficulty level assigning unit 17 includes a difficulty level dictionary generating unit 20 and a difficulty level determining unit 21. In the difficulty level dictionary generation unit 20, in addition to information for setting a basic ranking that is handled by a normal search engine, such as the frequency of occurrence of words and phrases in an electronic document collected from the Internet, the difficulty level of the electronic document The difficulty level dictionary 18 is generated / updated on the basis of information for determining the difficulty level, such as the feature amount related to the difficulty level (difficulty level dictionary generation step). As an example, when the difficulty level dictionary generation unit 20 makes a difficulty level assignment request, the language decoding level / difficulty level, such as the document creator's age, link status from an external site, etc., is collected from documents collected from the Internet or the like. Collecting documents that contain information that is closely related to, creating difficulty information from features, feature values such as expertise and word vectors estimated based on the type and frequency of technical terms, and a difficulty dictionary 18.

  In the difficulty level dictionary 18 generated / updated in this way, the appearance frequency of words / phrases corresponding to the difficulty level, the feature amount of the difficulty level, and the like are stored according to the level of the difficulty level. For example, if the difficulty S1 is set in stages according to age, the level of words and the frequency of appearance and difficulty according to the level are set for each level. Feature quantities and the like are stored.

  Note that the dictionary generation method is not limited to the above method as long as it can acquire information such as words and features according to the degree of difficulty. For example, a list of words that can be deciphered for each level of difficulty is generated from information of various dictionaries such as a kanji dictionary that includes information on the difficulty, and this may be stored in the difficulty dictionary 18. . Alternatively, a method of preliminarily selecting words and the like suitable for the decoding level of the user, and storing the words having high co-occurrence with those words as words belonging to the corresponding difficulty level in the difficulty dictionary It may be.

  The difficulty level determination unit 21 uses the difficulty level dictionary 18 to determine and assign a difficulty level Sl for each document to a document presented as a search result collected from the Internet or the like (difficulty level determination step). When the difficulty level Sl for each document is assigned in advance, a difficulty level index 19 database for difficulty level processing is prepared and stored in advance so that the difficulty level Sl can be referred to at the time of search execution. The difficulty index 19 may be in an appropriate form according to the processing speed or the like as long as the difficulty S1 can be extracted for each document. For example, the search engine stores search data for each document. It may be stored as a part of the index held in the unit 15. The processing step of the difficulty level assigning unit 17 ends with the generation of the index.

  The processing method of the difficulty level assigning unit 17 may be a method of sequentially calculating an electronic document searched by a search engine, in addition to the method previously given to each document as described above. In addition, each of the processing of the difficulty level dictionary generation unit 20 and the difficulty level determination unit 21 in the difficulty level assigning unit 17 may be sequentially calculated at the time of search in addition to the form performed in advance. Can be reflected, and there is an advantage that information freshness is excellent. Instead of re-sequentially processing all the documents, for example, only the document information included in the search result may be calculated and reflected. Furthermore, the amount of documents, the processing speed, and the required information Appropriate techniques can be employed, such as adjusting the processing content according to the freshness or the like, or changing the implementation opportunity. The search interface unit 24 will be described with reference to FIG. The document search apparatus 10 displays the search interface unit 24 on the browser of the user terminal 12 via the Internet. A user inputs a keyword group to be searched, that is, a query to the search interface unit 24 and requests a search process. At this time, it is estimated from the user's information (age, gender, etc.), the user's history (history, etc.), etc. that are input through the search interface 24 or in advance in a user DB (not shown), etc. Thus, the decryption level of the user is obtained (decoding level determination step). The decryption level corresponds to the difficulty level Sl set for each document described above, and is set to a plurality of levels (children, students, adults, etc.) according to the age, for example, as with the difficulty level Sl. The There is no need for such a combination of the decoding level and the difficulty level. For example, in addition to a combination based on age, a plurality of combinations such as a combination based on a degree of knowledge or experience in a predetermined specialized field may be used together. Also good. Note that the method for obtaining the user's decryption level is not limited to the above-described method. For example, as described in Patent Document 2, a search assistant is used to determine the user's decryption level at the time of retrieval. May be.

  When a user inputs a query through the user terminal 12 and requests a search, the query input from the search interface unit 24 and the decoding level of the user are passed to the ranking adjustment unit 23. This query is passed from the ranking adjustment unit 23 to the search engine unit 16.

  As shown in FIG. 3, the search engine unit 16 returns the IDs of a plurality of electronic documents including the query and the basic ranking score Sr of each electronic document to the ranking adjusting unit 23 as a search result for the query. (Search step). That is, the search engine unit 16 gives the basic ranking score Sr for each document in consideration of the relevance / matching degree (matching degree) to the query and the importance of the query in the document. Are ranked in descending order of the score Sr. That is, the higher the basic ranking score Sr, the higher the display order, and the higher priority is displayed. However, in the present embodiment, the ranking is adjusted through the ranking adjusting unit 23 according to the decoding level and the difficulty level of each document, and the adjusted ranking is returned and displayed to the user.

  Specifically, the ranking adjustment unit 23 refers to the difficulty index 19 when the search results (ID of a plurality of electronic documents including a query and basic ranking score Sr of each document) are returned from the search engine unit 16. Then, the difficulty level Sl is acquired for each document ID of the search result. Then, the ranking adjustment unit 23 obtains from the user decryption level obtained through the search interface unit 24 and the difficulty index 19 for each electronic document constituting the search result returned from the search engine unit 16. The basic ranking score Sr is adjusted according to the degree of difficulty, and a final ranking score Sri is obtained (ranking adjustment step).

  As an example of such ranking adjustment, for each of a plurality of electronic documents in the search result, a linear sum of the difficulty level Sl according to the user's decoding level and the document ranking score Si obtained from the search engine is calculated. . That is, an appropriate weight is added to the difficulty level Sl that matches the user's decoding level, and then added to the ranking score Si to calculate the final return ranking score Sri of each electronic document. Alternatively, the weight of the difficulty level Sl may be adjusted according to the degree of coincidence with the decryption level so that the score of the difficulty level Sl increases as the degree of coincidence with the decryption level of the user increases. By calculating these for all the documents acquired from the search engine, the final ranking score Sri of all searched electronic documents is obtained, and the search results are sorted in descending order of the ranking score Sri. The search results sorted here are transmitted to the user terminal 12 and displayed on the browser via the search interface unit 24.

  The final ranking score Sri is calculated by combining the electronic documents obtained from the search engine, the basic ranking score Sr, the difficulty S1 of these electronic documents, and the user's decoding level. If the condition of the method is satisfied, the method is not limited to that described above.

  As described above, in the present embodiment, the electronic documents having the difficulty level corresponding to the level of decoding such as the user's experience, knowledge and language level are preferentially ranked for the search results included in the search according to the query. By returning and presenting it to the user in this way, the efficiency with which the user can decode and understand the information can be greatly improved.

  In addition, since the ranking is adjusted according to the user's decoding level, some electronic documents are not excluded from the search results as in the filter processing described above. For this reason, even if the electronic document with the difficulty level that matches the decryption level is preferentially placed at the top of the ranking, the ranking of the electronic document containing useful information may be lowered, but the user can By sequentially referring to the document, it is possible to reliably access the electronic document including this useful information. Furthermore, by creating and updating the difficulty level dictionary in advance, it becomes possible to determine the difficulty level of each electronic document more accurately and quickly.

  Note that the present invention can also be configured as a program that causes a computer to function as a part or all of the processing contents of the units 13, 14, 16, 17, 20, 21, 21, 23, and 24 of the document search apparatus 10 described above. . With this program, a part or all of the processing contents can be executed by a computer.

  This program can be provided through a network such as a website or e-mail. The program is recorded on a recording medium such as a CD-ROM, DVD-ROM, CD-R, CD-RW, DVD-R, DVD-RW, MO, HDD, Blu-ray Disk (registered trademark). It is also possible to save and distribute. This recording medium is read using a recording medium driving device, and the program code itself realizes the processing of the above embodiment, so that the recording medium also constitutes the present invention.

DESCRIPTION OF SYMBOLS 10 ... Document search device 12 ... User terminal 16 ... Search engine part (search means)
17 ... Difficulty level giving unit (difficulty level giving means)
18 ... Difficulty level dictionary 21 ... Difficulty level determination unit (difficulty level determination means)
23 ... Ranking adjustment section (ranking adjustment means)
24 ... Search interface section (decoding level determination means)

Claims (5)

Search means for searching for an electronic document group based on a query input by a user, and setting a ranking according to the degree of relevance to the query for the electronic document included in the search result;
A decryption level judging means for judging a decryption level of the user;
Difficulty level giving means for giving a difficulty level related to the decoding level to at least an electronic document included in the search result;
Ranking adjustment means for adjusting the ranking according to the difficulty level and the decoding level, and preferentially positioning the electronic document with the difficulty level according to the decoding level at the top of the ranking;
A document search apparatus characterized by comprising: The difficulty level giving means is:
Difficulty level dictionary generating means for acquiring information related to the difficulty level from the electronic document group, and generating a difficulty level dictionary based on the information related to the difficulty level;
A difficulty level determination means for determining the difficulty level of each electronic document with reference to the difficulty level dictionary;
The document search apparatus according to claim 1, further comprising: A search step in which the search means searches the electronic document group based on the query input by the user, and sets a ranking according to the degree of relevance to the query for the electronic document included in the search result;
A decryption level determination means for determining a decryption level of the user;
A difficulty level assigning step in which the difficulty level assigning means assigns a difficulty level related to the decoding level to at least an electronic document included in the search result;
A ranking adjustment step, wherein the ranking adjustment means adjusts the ranking according to the difficulty level and the decoding level, and ranks the electronic document with the difficulty level according to the decoding level preferentially positioned at the top of ranking
A document search method characterized by comprising: The difficulty level assigning step includes:
A difficulty level dictionary generating step of acquiring information related to the difficulty level from the electronic document group and generating a difficulty level dictionary based on the information related to the difficulty level;
A difficulty level determination step of determining the difficulty level of each electronic document with reference to the difficulty level dictionary;
The document search method according to claim 3, further comprising:   A document search program for causing a computer to function as each of the means constituting the document search device according to claim 1.

Images