JP2014067374A - Index construction device for search, index construction method for search and program for index construction device for search - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow, even after a Web page is created, a keyword to be associated with the Web page to function as a related keyword of the Web page even when the number of the keywords is increased.SOLUTION: Whether or not a musical piece name is included in an answer column is determined by performing access to a Q & A page to refer to the HTML data of the Q & A page. When such a musical piece name is included, a keyword is extracted from a question column, and stored so as to be associated with the musical piece name in a musical piece DB 4. The musical piece name is acquired by performing access to a musical piece introduction page, and whether or not the keyword associated with the musical piece name is registered in the musical piece DB 4 is inquired. When the keyword is registered, the keyword is obtained, and embedded in the header part of the HTML data of the Q & A page.

Description

  The present invention relates to a technique for constructing a search index.

When searching for a web page on the Internet, a technique for using a part of data of a web page as a search keyword to improve the search efficiency is well known (for example, Patent Document 1).
Specifically, a keyword related to a web page is embedded in a non-displayed data area called a header in HTML (HyperText Markup Language) data. The search engine system periodically circulates web pages that are open to the public on the Internet, collects the keywords embedded in the header, associates the web pages with the collected keywords, and creates a database.

When a user of the Internet inputs a search word using a search engine, the search engine determines whether or not the input search word matches the keyword stored in the database. As a web page related to the input search word, the link destination of the web page stored in the database in association with the keyword is displayed as a search result.
Therefore, by appropriately adopting the keyword embedded in the header, it is possible to easily display a specific web page as a search result.

  Taking a web page that introduces a song as an example, by embedding keywords in the header such as a song name and artist name, and keywords summarizing the text of the web page that introduces the song, for example, a user When searching for the song name as a search word, a web page introducing the song can be displayed as a search result.

JP 2011-54102 A

  However, since the keyword embedded in the header is a keyword adopted at the time of creating the web page, it may not be compatible with the latest search word used by Internet users.

For example, it is assumed that when a certain music is released, a web page that introduces the music is created, and after the release, the music is used in a certain CM. In this case, when a search is performed using a word related to the CM as a search word, it is desirable that a web page introducing the music is displayed as a search result.
However, since the word related to the CM is not embedded in the header when the web page introducing the music is created, the music is introduced as a search result even if a search is performed using the word related to the CM. Web pages to be displayed are not displayed.
The present invention has been made in view of this point, and even when the number of keywords to be associated with the web page increases after the creation of the web page, the keyword is used as a keyword for searching the web page. Make it function as a solution issue.

  Means employed by the present invention to solve the above problems will be described below. In order to facilitate understanding of the present invention, reference numerals in the drawings will be appended in parentheses for convenience in the following, but the present invention is not intended to be limited to the illustrated forms.

    In order to solve the above-described problem, the search index construction device according to the present invention is a search index construction device (1) that constructs a search index used for searching the first document by the search device (3). A determination unit (10) for determining whether or not the second document different from the first document includes a predetermined main keyword stored in association with the first document; and the determination unit (10 ) To extract a keyword corresponding to the main keyword from the second document determined to contain the main keyword, and to extract the keyword extracted by the extraction unit (11) to the first document And an update unit (12) for updating the search index as a related keyword.

  In the present invention, the search index may be used by a general-purpose search device, or may be used by a dedicated search device. A general-purpose search device is a search device that targets all web pages searchable on the Internet, for example, and a dedicated search device is a specific web as specified in the embodiment of the present invention, for example. This search device targets only a page (music introduction page). Details of the dedicated search device will be described later. As long as the search index includes a related keyword group, the place to be provided may be inside the first document, an external database, or the like. When the search index is provided in the first document, it may be not only the header portion of the first document but also other portions. The first document may have any content as long as it is a document to be displayed as a search result, and the main keyword may be any keyword as long as it is a keyword mainly related to the first document.

  In the above-described search index construction device (1), the update unit (12) uses the keyword extracted by the extraction unit (11) as the related keyword of the first document and uses the search index in the first document. You may make it update.

  In the above-described search index construction device (1), the first document and the second document are web pages described in a markup language, and the updating unit (12) is configured by the extracting unit (11). You may make it update the said index for a search in the header of the said web page by making the extracted keyword into a related keyword of the said 1st document.

  In the search index construction device (1) described above, the updating unit (12) stores the keyword extracted by the extraction unit (11) as a related keyword of the first document and stored in the database (4). The index may be updated.

  In the search index construction device (1) described above, the extraction unit (11) may extract proper nouns as keywords.

  In the search index construction device (1) described above, a question page on which an answer to a question can be posted is used as the second document, and a main keyword is included in the answer portion of the question page by the determination unit (10). If it is determined, the extraction unit (11) extracts a keyword corresponding to the main keyword from the question part, and the determination unit determines that the main keyword is included in the question part of the question page. In such a case, the extraction unit may extract a keyword corresponding to the main keyword from the answer portion.

    In order to solve the above-described problems, a search index construction method according to the present invention is a search index construction method for constructing a search index used for searching a first document by a search device, wherein the first document The second document different from the first document determines whether or not the predetermined main keyword stored in association with the first document is included, and the determination unit determines that the main keyword is included. A keyword corresponding to the main keyword is extracted from two documents, and the search index is updated using the extracted keyword as a related keyword of the first document.

  In order to solve the above-described problem, a search index construction device program according to the present invention constructs a search index construction device (1) that constructs a search index used for retrieval of a first document by a retrieval device (3). Whether or not a predetermined main keyword stored in association with the first document is included in a second document different from the first document. A determination unit (10) for determining, an extraction unit (11) for extracting a keyword corresponding to the main keyword from the second document determined by the determination unit (10) to include the main keyword; The search index is updated with the keyword extracted by the extraction unit (11) as the related keyword of the first document. Characterized in that to function as a new part (12).

  The program may be stored in a computer-readable recording medium. If this recording medium is used, the program can be installed in the computer, for example. Here, the recording medium storing the program may be a non-transitory recording medium such as a CD-ROM.

It is a block diagram for demonstrating the environment where the index construction | assembly apparatus for a search which concerns on 1st Embodiment of this invention is used. It is explanatory drawing which shows the structure of the data memorize | stored in music DB. It is a block diagram which shows the structure of the index construction | assembly apparatus for a search. It is a flowchart which shows operation | movement of the index construction | assembly apparatus for a search. It is a flowchart which shows operation | movement of the index construction | assembly apparatus for a search. It is explanatory drawing which shows an example of a Q & A page. It is explanatory drawing which shows an example of a part of HTML data of a music introduction page. It is a block diagram for demonstrating the environment where the index construction | assembly apparatus for a search which concerns on 2nd Embodiment of this invention is used. It is explanatory drawing which shows the structure of the data memorize | stored in music DB of 2nd Embodiment. It is a flowchart which shows operation | movement of the index construction apparatus for a search in a modification.

Hereinafter, as an embodiment, a search index construction device according to the present invention will be described with reference to the drawings.
<First Embodiment>
<1. Overview of Search Index Building Device>
FIG. 1 is a block diagram for explaining an environment in which the search index construction device 1 according to the first embodiment of the present invention is used. Here, the search index is information used for searching the first document by the search device 3, and is composed of a main keyword and a related keyword. The main keyword is a keyword representing the subject of the first document as the search target, and the related keyword is a keyword related to the content of the first document as the search target. Details of the search index will be described later. As shown in FIG. 1, the search index construction device 1 can communicate with each of a music DB (database) 4, a public server 5, and a hosting server 6 via a communication network NET such as the Internet. . The user terminal 2 can communicate with each of the general-purpose search device 3, the public server 5, and the hosting server 6 via the communication network NET. Furthermore, the general-purpose search device 3 can communicate with each of the public server 5 and the hosting server 6 via the communication network NET.

  The user terminal 2 can communicate via the communication network NET, and corresponds to, for example, a personal computer, a mobile phone, a smartphone, a tablet terminal, or the like. The user terminal 2 is installed with browser software capable of displaying a web page published on the communication network NET. By using the browser software, HTML data or the like is displayed on the user terminal 2 as a web page. Can be displayed. Moreover, character data etc. can be input into the input column etc. of a web page using the keyboard etc. with which the user terminal 2 was equipped. Although only one user terminal 2 is shown in FIG. 1, a plurality of user terminals 2 are connected to the communication network NET.

  The general-purpose search device 3 periodically circulates through the public server 5 and the hosting server 6 and accesses web pages published on these servers, so that the search device 3 can access the web page from the HTML data of the web page. Collect keywords to use when searching web pages. Then, the web page and the collected keyword are associated and recorded in a DB (database) 3a. In addition, the search device 3 is provided with HTML data of a search web page as a user interface that allows a user to input a search word for searching the web page, and so on. The search word can be input by accessing the search web page using the browser software 2. When a search word is input in the input field of the search web page and a button for instructing execution of the search is pressed, the search device 3 matches the keyword recorded in the DB 3a with the input search word. If they match, the link destination of the web page associated with the keyword is displayed as a search result on the search web page. Therefore, the user can access the web page related to the search word.

The public server 5 is a server that can be accessed via the communication network NET, and includes an HDD (Hard Disk Drive) 5a capable of recording programs and data. In the present embodiment, as an example, it is assumed that HTML data of a music introduction page (an example of a first document) that is a web page for introducing music is recorded in the HDD 5a. Therefore, the user can obtain information related to the music piece by accessing the music introduction page using the browser software of the user terminal 2.
Although only one public server 5 is shown in FIG. 1, HTML data of the music introduction page may be stored in a plurality of public servers 5. The public server 5 can be replaced with a hosting server.

  The hosting server 6 is a server that can be accessed via the communication network NET, and includes an HDD (Hard Disk Drive) 6a capable of recording programs and data. In this embodiment, as an example, it is assumed that HTML data of a Q & A page (an example of a second document) is recorded in the HDD 6a. The Q & A page is a web page used for a service in which when a certain user posts a question, an answer to the question can be entered by another user. The Q & A page does not indicate only one page in a specific service, but may be in any form as long as the answer to the question is posted. There are multiple types of Q & A pages. doing. These plural types of Q & A pages may be stored in one hosting server 6 or may be stored in a plurality of hosting servers 6. Although only one hosting server 6 is shown in FIG. 1, a plurality of hosting servers 6 may be used. The hosting server 6 can be replaced with a public server.

  With reference to FIG. 1, functions of the search index construction device 1 will be described. The search index construction device 1 includes a determination unit 10, an extraction unit 11, an update unit 12, and a storage unit 13. The determination unit 10, the extraction unit 11, and the update unit 12 are functional blocks that are exhibited when the search index construction device 1 executes a program for construction of a search index.

The determination unit 10 determines whether or not the HTML data of the Q & A page includes a predetermined main keyword (music name, etc.) stored in association with the reference information (file name or location information such as URL) of the music introduction page. Judging.
When the determination unit 10 determines that the main keyword is included in the answer column on the Q & A page, the extraction unit 11 extracts a keyword corresponding to the main keyword from the question column on the Q & A page. The keyword corresponding to the main keyword is a keyword described in the question column in relation to the main keyword. Details will be described later.

  The update unit 12 updates the search index with the keyword extracted by the extraction unit 11 as the related keyword of the music introduction page. In this embodiment, as an example, the search index is provided in the header of HTML data of the music introduction page. The search device 3 accesses each web page and collects keywords related to each web page from the HTML data of the web page by various methods. As one of them, the header of the HTML data of each web page This is done by collecting words included in the search index provided in the folder. Therefore, by updating the search index with the keyword extracted by the extraction unit 11 as the related keyword of the music introduction page, the search device 3 searches for the extracted keyword as a search word. An introduction page is expected to be displayed.

  The storage unit 13 stores a program for controlling the search index construction device 1, and further stores URLs of one or more Q & A pages. Therefore, the search index construction apparatus 1 accesses the Q & A page with reference to the URL of the Q & A page stored in this way, and associates the Q & A page with the music introduction page by the determination unit 10 as described above. It is determined whether or not the predetermined main keyword (music name, etc.) stored is included.

The music DB 4 can communicate with the search index construction device 1 via the communication network NET. As shown in FIG. 2, the music DB 4 is associated with a main keyword (music name, etc.) and a related keyword. It is remembered. In this embodiment, a plurality of music introduction pages are prepared in advance, and the music names of the music introduced on those music introduction pages are stored in the music DB 4 as main keywords. The search index construction device 1 refers to the music DB 4 as needed to determine the presence or absence of the main keyword on the Q & A page described above or update the search index corresponding to the music introduction page. Details will be described later.
In this embodiment, the example in which the music DB 4 is configured separately from the search index construction device 1 has been described. However, the music DB 4 is provided as one of the components of the search index construction device 1. May be.

FIG. 3 shows the configuration of the search index construction device 1. As shown in this figure, a search index construction device 1 includes a CPU (Central Processing Unit) 30 that controls the entire device, a RAM (Random Access Memory) 31 that functions as a work area of the CPU 30, a ROM that stores a boot program, and the like. (Read Only Memory) 32, a hard disk drive (HDD) 33 for storing various programs and data, an input unit 34 including a keyboard and a mouse, a display 35 for displaying images, and communication with an external device via a communication network NET And a reader 37 for reading an information recording medium such as a compact disk. The HDD 33 is an example of the storage unit 13 described above.
In the present embodiment, the CPU 30 can operate as the determination unit 10, the extraction unit 11, and the update unit 12.

<2. Operation of Search Index Building Device>
Hereinafter, the operation of the search index construction device 1 of the present embodiment will be described with reference to FIGS.

<2-1: Keyword extraction processing>
First, the keyword extraction process will be described with reference to the flowchart of FIG. The CPU 30 of the search index construction device 1 inquires of the music DB 4 what kind of music name is registered (S200). The music DB 4 that has received the inquiry request from the CPU 30 of the search index construction device 1 extracts only the music name from the registered information and creates a music list (S300). In the case of the present embodiment, as shown in FIG. 2, “DDD”, “EEE”, “FFF”, and the like are stored as song names, and a song list is created by extracting these song names. When the creation of the music list is completed, the music DB 4 notifies the search index construction device 1 of the created music list (S301).

  When the CPU 30 of the search index construction device 1 acquires the music list from the music DB 4 (S201), the CPU 30 accesses the Q & A page by referring to the URL of the Q & A page stored in the storage unit 13, and stores the Q & A page in the HTML data. The date data included is confirmed (S202). The date data indicates the date when the Q & A page was last updated. The storage unit 13 of the search index construction device 1 stores date data acquired when the Q & A page was last accessed. The CPU 30 of the search index construction device 1 compares the stored date data with the date data included in the HTML data corresponding to the Q & A page accessed this time, and the Q & A page is more recently accessed than the previous access. It is determined whether it has been updated later (S203). If it is determined that it has not been updated after the previous access (S203; NO), the process proceeds to another process.

  However, if it is determined that the Q & A page has been updated after the previous access (S201; YES), the CPU 30 of the search index construction device 1 stores the document text data included in the HTML data of the Q & A page. Obtain (S204). Then, the part corresponding to the answer column is referred to from the document text data, and the music title is searched from the part (S205).

  FIG. 6 shows an example of the Q & A page. As shown in FIG. 6, the Q & A page P1 includes a question column A1 and an answer column A2, and when a user enters a question in a question column (not shown), the question is entered in the question column A1. It is displayed. When another user who has viewed the question displayed in the question column A1 enters an answer to the question in the answer entry column (not shown), the answer is displayed in the answer column A2. . In the example shown in FIG. 6, text data “Q:” is described in the title portion of the question column A1, and text data “A:” is described in the title portion of the answer column A2. Therefore, by searching for text data “A:” from the document text data on the Q & A page, it is possible to refer to the contents of the answer described in the answer column A2. In Fig. 6, the user with the user name “XXX” wants to know the song used in the CM of the BBB (car name) of Company A. If anyone knows it, please let me know. Please post the question "Please." In addition, in this example, another user having a user name “ΔΔΔ” posted an answer “DCC (music name) of CCC (artist name)”.

  To extract a song name from a sentence, a natural language processing technique called morphological analysis, a well-known algorithm such as a TD-IF method, or the like is used. Specifically, a proper noun is extracted from a sentence. In the example shown in FIG. 6, the artist name “CCC” and the song name “DDD” are extracted as proper nouns as search results.

  The CPU 30 of the search index construction device 1 compares the extracted proper noun with the music list acquired in step S201, and determines whether or not the music name is included in the extracted proper noun (S206). . If it is determined that the song name is not included (S206: NO), the process proceeds to another process. However, if it is determined that the song name is included (S206: YES), the CPU 30 of the search index construction device 1 extracts a keyword from the question column A1 (S207). In the example described above, the extracted proper nouns are “CCC” and “DDD”, but among these, “DDD” is included in the song list as the main keyword (music name), so the search index construction device 1 CPU30 will judge that the music name was contained in the extracted proper noun.

  In order to extract a keyword from the question column A1, as in the case of searching for a song name from a document text, a natural language processing technique called morphological analysis or a well-known algorithm such as a TD-IF method is used. Proper nouns are extracted. In the example shown in FIG. 6, the question column A1 states “I want to know the song used in the CM of A company's BBB (car name), but please let me know if you know it”. Therefore, “CM of BBB (car name)” is extracted as a keyword.

  When the CPU 30 of the search index construction device 1 extracts a keyword, the CPU 30 notifies the music DB 4 of the extracted keyword (S208). The music DB 4 that has received the notification determines whether or not the keyword is a new keyword, that is, whether or not the keyword is already registered (S302). Specifically, it is determined whether or not the notified keyword is included in the related keywords stored in association with the registered main keyword (music name). As a result, if it is a new keyword (S302: YES), it is registered as a related keyword (S303), and the result is notified to the search index construction device 1 (S304). FIG. 2 shows an example in which the keyword “CM of BBB (car name)” is registered in association with the song name “DDD”. However, if it is not a new keyword, registration processing is not performed, and the result is notified to the search index construction device 1 (S304). Then, the CPU 30 of the search index construction device 1 acquires the registration result from the music DB 4 (S209).

<2-2: Music introduction page update process>
Next, the music introduction page update process of this embodiment will be described with reference to FIG. The music introduction page update process may be performed continuously after the keyword acquisition process shown in FIG. 4 ends, or may be performed independently of the keyword acquisition process. In the present embodiment, an example will be described which is performed after the keyword acquisition process illustrated in FIG. 4 is completed.
The URL of the music introduction page is stored in the storage unit 13 of the search index construction apparatus 1 in advance, and the CPU 30 of the search index construction apparatus 1 refers to the URL of the music introduction page to thereby refer to the music introduction page. To access. Then, the document text data of the music introduction page is acquired (S221).

  The CPU 30 of the search index construction device 1 acquires the music title from the acquired document text data (S222). FIG. 7 shows an example of HTML data of the music introduction page. As shown in FIG. 7, the HTML data P2 of the music introduction page includes a header part B1 surrounded by tags <head> and </ head>, and a text surrounded by tags <body> and </ body>. Part B2. Note that the above-described search index corresponds to the header portion B1. Since the music introduction page prepared in advance in the present embodiment is titled “DDD (song title) introduction page” at the beginning of the body part, the part before the “introduction page” in this title is read. The song name is acquired.

  The CPU 30 of the search index construction device 1 inquires of the music DB 4 whether or not the keyword for the acquired music name is registered (S223). The music DB 4 that has received the inquiry searches whether or not the related keyword corresponding to the inquired music name is registered (S310). Then, the search result is notified to the search index construction device 1 (S311), and the CPU 30 of the search index construction device 1 acquires the search result (S224). In the above-described example, as shown in FIG. 2, since the related keyword “CM of BBB (car name)” is registered in association with the music name “DDD”, the music DB 4 stores “BBB (car name)”. The search index construction apparatus 1 is notified that the related keyword “CM of“ is registered ”.

  The CPU 30 of the search index construction device 1 determines whether or not a related keyword is registered in the music DB 4 based on the acquired search result (S225). In the above example, since the related keyword “CM of BBB (car name)” is registered, the related keyword of “CM of BBB (car name)” is surrounded by the <head> and </ head> tags. The embedded header is embedded (S226). More specifically, as shown in FIG. 7, the header portion includes a portion where <meta name = "keywords" content = "keyword 1, keyword 2, keyword 3">. The related keywords notified from the music DB 4 are embedded in the “keyword 2, keyword 3” portion. For example, it is assumed that when a music introduction page is created, a music name is embedded as keyword 1 and an artist name is embedded as keyword 2. In this way, at the time of creating the music introduction page, with the two keywords embedded, the update process of the music introduction page of the present embodiment causes the keyword 3 to be replaced with “BBB (car name) as the third keyword. Embedded CM ”. In this way, a new keyword is extracted from the question column of the Q & A page by the keyword extraction process shown in FIG. 4, and the new extracted part is added to the header portion of the music introduction page by the update process of the music introduction page shown in FIG. Keywords are added.

  Therefore, at the beginning of the creation of the music introduction page, when a user searches using the search device 3 with the music name or artist name as a search word, the music name and artist name are Since it is embedded in the header portion of the introduction page, this music introduction page is displayed as a search result. At this point, even if a search is performed using the search word “BBB (car name) CM”, this search word is not yet embedded in the header portion of the music introduction page. The music introduction page is never displayed.

  However, in the present embodiment, as described above, a new keyword “CM of BBB (car name)” is added to the header portion of the music introduction page by referring to the Q & A page. When searching with the search word “BBB (car name) CM” using the device 3, the music introduction page is displayed as a search result.

  As described above, according to this embodiment, when a song name is included in the answer column of the Q & A page, a new keyword is extracted from the question column, and the song introduction page, which is a page for introducing the song, is displayed. Since it is added as a related keyword in the header part, even if there are more keywords to be associated with the music introduction page after creating the music introduction page, the keyword is sent to the general-purpose search device. Can effectively function as a related keyword.

Second Embodiment
Next, a second embodiment of the present invention will be described with reference to FIGS. In the first embodiment, an example in which a general-purpose search device is used as the search device has been described. However, in the second embodiment, as illustrated in FIG. 8, a configuration using a dedicated search device 3A for a music introduction page is used. Different from one embodiment.

  Further, as shown in FIG. 9, the music DB 4A in the second embodiment not only stores related keywords in association with main keywords (music names), but also references information (URL) on the music introduction page. It is stored in association with the name.

  Also in the second embodiment, the keyword extraction process shown in FIG. 4 is performed in the same manner as in the first embodiment, but in the second embodiment, the music introduction page update process shown in FIG. 5 is not performed. The search index update process is only a process of updating the related keywords in the music DB 4A in the keyword extraction process shown in FIG. That is, in the first embodiment, the example in which the header portion of the HTML data of the music introduction page functions as a search index has been described, but in the second embodiment, the music DB 4A itself functions as a search index. ing.

  In the second embodiment, when a certain user performs a search using “CM for BBB (car name)” using the dedicated search device 3A as a search word, the search device 3A searches for “BBB (car name)”. It is determined whether or not “CM” is registered as a related keyword in the music DB 4A. If it is registered as a related keyword, the URL of the music introduction page stored in association with the related keyword is displayed as a search result.

  Therefore, also in this embodiment, when a song name is included in the answer column of the Q & A page, a new keyword is extracted from the question column, and a song introduction page that is a page for introducing the song in the song DB is displayed. Since it is added as a related keyword in association with the URL, even if the number of keywords to be associated with the music introduction page increases after the creation of the music introduction page, the keyword is assigned to the dedicated search device described above. It can function effectively as a related keyword of the music introduction page.

<Modification>
The present invention is not limited to the above-described embodiments, and various modifications described below are possible. Of course, each modification and embodiment may be appropriately combined.

  In the above-described embodiment, the example has been described in which the music DB 4 and the music DB 4A are separated from the search index construction device 1 and can communicate with the search index construction device 1 via the communication network NET. Is not limited to such an example. That is, the search index construction device 1 may have the functions of the music DB 4 and the music DB 4A. Alternatively, information stored in the music DB 4 and the music DB 4A may be stored in the storage unit 13 of the search index construction device 1 as a table, and the search index construction device 1 may search the table. In addition, the individual elements in the music DB 4 and music DB 4A information stored as a table are stored in a plurality of different storage devices or a plurality of different storage areas in the same storage device as long as they are associated with each other so that they can be referred to each other. It may be stored in a separated state.

  In the first embodiment, the example in which the header portion of the HTML data is updated as the search index has been described. However, depending on the general-purpose search device, the related keyword is extracted by referring to a portion other than the header portion of the HTML data. Therefore, when it is assumed that such a general-purpose search device is used, a part other than the header part of the HTML data may be updated as a search index.

  In the above-described embodiment, the example in which the song name is extracted from the answer column of the Q & A page and the keyword is extracted from the question column has been described. However, for example, a question about what CM is used for a predetermined song Since a predetermined CM name may be posted as an answer, in such a case, as shown in FIG. 10, the song name is extracted from the question column on the Q & A page (S230), and the answer column is displayed. A keyword may be extracted (S231).

  In the above-described embodiment, an example in which a Q & A page is used as a target page from which a new keyword is extracted has been described. However, the present invention is not limited to such an example, and is displayed as a search result. For example, a page such as a blog may be used as long as it is a page related to the main keyword of the page to be processed. In other words, in the above-described embodiment, the page to be displayed as a search result is the music introduction page, and the main keyword of the music introduction page is the music name, so if it is a page related to the music name, other than the Q & A page Blog pages, news related pages, etc.

  Furthermore, the page to be displayed as the search result is not limited to the music introduction page, and may be any page such as a page for introducing a product. The concept of a page is not limited to a web page formed as HTML data. As long as it is possible to use the same method as in the above-described embodiment, the search target by the user and the keyword extraction target may be, for example, a text file or a document file that can be read by an arbitrary application.

  Note that a program for realizing the functions of the present invention may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read by a computer system and executed. Here, the “computer system” includes an OS and hardware such as peripheral devices. Further, the “computer system” may include a plurality of computer devices connected via a network including a communication line such as the Internet, WAN, LAN, and dedicated line. The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Furthermore, the “computer-readable recording medium” holds a program for a certain period of time, such as a volatile memory (RAM) inside a computer system that becomes a server or a client when the program is transmitted via a network. Including things. The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient. Also, a distribution server that distributes a program for realizing the function of the present invention or a part thereof, a storage medium provided in the distribution server, and a distribution server that exists outside the distribution server and distributes the program by the distribution server A storage medium stored for this purpose is also included in the scope of the present invention.

  In addition, some or all of the functions described above may be realized as an integrated circuit such as an LSI (Large Scale Integration). Each function described above may be individually made into a processor, or a part or all of them may be integrated into a processor. Further, the method of circuit integration is not limited to LSI, and may be realized by a dedicated circuit or a general-purpose processor. In addition, when an integrated circuit technology that replaces LSI appears due to the advancement of semiconductor technology, an integrated circuit based on the technology may be used.

  In addition, this invention is not limited to the above-mentioned embodiment and modification, The change within the range of the meaning of this invention is included in this invention.

  NET ... Communication network, 1 ... Index building device for search, 2 ... User terminal, 3,3A ... Search device, 4,4A ... Music DB, 5 ... Public server, 6 ... Hosting server, 10... Judging unit, 11... Extracting unit, 12... Updating unit, 13.

Second Embodiment
Next, a second embodiment of the present invention will be described with reference to FIGS. In the first embodiment, an example in which a general-purpose search device is used as the search device has been described. However, in the second embodiment, as shown in FIG. 8, a dedicated music introduction page having a DB (database) 3b is used. The configuration using the search device 3A is different from that of the first embodiment.

Claims (8)

A search index building device for building a search index used for searching a first document by a search device,
A second document different from the first document includes a determination unit that determines whether or not a predetermined main keyword stored in association with the first document is included, and the determination unit includes the main keyword. An extraction unit for extracting a keyword corresponding to the main keyword from the second document determined to be, and an update unit for updating the search index with the keyword extracted by the extraction unit as a related keyword of the first document; Comprising
An index construction device for search.   2. The search index construction device according to claim 1, wherein the update unit updates the search index in the first document using the keyword extracted by the extraction unit as a related keyword of the first document. .   The first document and the second document are web pages described in a markup language, and the update unit uses the keyword extracted by the extraction unit as a related keyword of the first document in the header of the web page. The search index construction apparatus according to claim 2, wherein the search index is updated.   The update unit updates the search index stored in a database using the keyword extracted by the extraction unit as a related keyword of the first document. The index construction device for search described.   The search index construction apparatus according to any one of claims 1 to 4, wherein the extraction unit extracts proper nouns as keywords.   As the second document, a question page on which an answer to a question can be posted is used, and when the determination unit determines that a main keyword is included in an answer part of the question page, the extraction unit When a keyword corresponding to a keyword is extracted from a question portion and the determination unit determines that a main keyword is included in the question portion of the question page, the extraction unit selects a keyword corresponding to the main keyword. 6. The search index construction apparatus according to claim 1, wherein the retrieval index construction apparatus extracts the answer part. Method corresponding to claim 1 Claim A search index construction method for constructing a search index used for searching a first document by a search device,
Determining whether the second document different from the first document includes a predetermined main keyword stored in association with the first document;
Extracting a keyword corresponding to the main keyword from the second document determined by the determination unit to include the main keyword;
Updating the search index using the extracted keyword as a related keyword of the first document;
A search index construction method characterized by that. A program claim corresponding to claim 1 is a program for a search index construction device comprising a computer for constructing a search index used for searching a first document by a search device,
The computer,
A determination unit for determining whether or not a second main document different from the first document includes a predetermined main keyword stored in association with the first document;
An extraction unit that extracts a keyword corresponding to the main keyword from the second document that is determined by the determination unit to include the main keyword;
Causing the keyword extracted by the extraction unit to function as an update unit that updates the search index as a related keyword of the first document;
A search index construction apparatus program characterized by the above.

Images