JP6079207B2 - Keyword presentation program, keyword presentation method, and keyword presentation apparatus - Google Patents

Description

  The present invention relates to a keyword presentation program, a keyword presentation method, and a keyword presentation device.

  A keyword search for searching for target data using a keyword as a clue is known. For example, the keyword search is used when a Web search for searching a Web page or a file such as an image or a sound in the Web page or a desktop search for searching a file stored on a computer is performed.

  When keywords used for such keyword search are general, a huge number of search results may be hit. For this reason, an AND search that searches for target data using a plurality of keywords specified by the user may be used. In such AND search, keyword input is also assisted by presenting words that are highly relevant to the input keyword or words that have a strong co-occurrence relationship as candidates.

Japanese Patent Laid-Open No. 10-260990 JP-A-5-054082 JP 2003-108594 A

  However, the above-described keyword input assistance has a problem that the search results cannot be narrowed down accurately as described below.

  For example, when a highly relevant word is presented with the above input assistance, a file including synonyms and synonyms of the input keyword is searched. For this reason, there are many cases where only a search result that is not much different from the case of searching by the input keyword alone is obtained, and it is not always possible to narrow down the search result. In addition, when a word having a strong co-occurrence relationship is presented with the above input assistance, a file further including a word that frequently appears together with the input keyword is searched. However, even when words with strong co-occurrence relationships are presented, there are many cases where words that are highly relevant to the entered keyword are presented, and only the search results that are not much different from the case of searching with the entered keyword alone are displayed. In many cases, it cannot be obtained.

  In one aspect, an object of the present invention is to provide a keyword presentation program, a keyword presentation method, and a keyword presentation device that can improve the accuracy of narrowing search results.

  The keyword presentation program according to one aspect causes a computer to execute processing for accepting specification of a keyword. Furthermore, the computer includes a plurality of files or a first storage unit that stores an index of a character string included in each of the plurality of files, or included in the metadata of the plurality of files or the metadata of each of the plurality of files. The second storage unit that stores the index of the character string to be stored is referred to, and a process for searching for a file that hits the specified keyword is executed. Further, the computer is caused to execute a process of creating an appearance map by setting the presence or absence of the specified keyword for each file or file metadata in accordance with a file search result. Furthermore, using the appearance map of the specified keyword and the appearance map of the keyword combined with the specified keyword, the computer extracts keywords that achieve a predetermined number of search results for the combination with the specified keyword. To execute the process. Further, the computer is caused to execute processing for determining the validity of the extracted keyword combined with the designated keyword. Further, the computer is caused to execute a process of presenting a keyword determined to be valid.

  According to one embodiment, it is possible to improve the search result narrowing accuracy.

FIG. 1 is a diagram illustrating the configuration of the search system according to the first embodiment. FIG. 2 is a block diagram illustrating a functional configuration of the server apparatus according to the first embodiment. FIG. 3 is a diagram illustrating an example of an appearance map of a designated keyword. FIG. 4 is a diagram illustrating an example of an AND operation. FIG. 5 is a diagram illustrating an example of an AND operation. FIG. 6 is a diagram illustrating an example of a keyword route. FIG. 7 is a diagram illustrating an example of the relationship between the number of search results for a specified keyword and the number of keywords combined with the specified keyword. FIG. 8 is a diagram illustrating an example of screen transition displayed on the client terminal. FIG. 9 is a diagram illustrating an example of screen transition displayed on the client terminal. FIG. 10 is a flowchart illustrating the procedure of the search process according to the first embodiment. FIG. 11 is a flowchart (1) illustrating the procedure of the process for generating the presented keyword route according to the first embodiment. FIG. 12 is a flowchart (2) illustrating the procedure of the process for generating the presented keyword route according to the first embodiment. FIG. 13 is a schematic diagram illustrating an example of a computer that executes a keyword presentation program according to the first and second embodiments.

  Hereinafter, a keyword presentation program, a keyword presentation method, and a keyword presentation apparatus according to the present application will be described with reference to the accompanying drawings. Note that this embodiment does not limit the disclosed technology. Each embodiment can be appropriately combined within a range in which processing contents are not contradictory.

[System configuration]
FIG. 1 is a diagram illustrating the configuration of the search system according to the first embodiment. A search system 1 shown in FIG. 1 provides a Web search service for searching for a Web page or a file such as an image or a sound in a Web page via a network 5 to client terminals 30A to 30C.

  As illustrated in FIG. 1, the search system 1 accommodates a server device 10 and client terminals 30 </ b> A to 30 </ b> C. Although FIG. 1 illustrates three client terminals, the search system 1 can accommodate any number of client terminals. In the following, when the devices of the client terminals 30A to 30C are collectively referred to without distinction, they may be referred to as “client terminals 30”.

  The server device 10 and the client terminal 30 are connected via the network 5 so that they can communicate with each other. The network 5 can employ any type of communication network such as the Internet (Internet), LAN (Local Area Network), VPN (Virtual Private Network), etc., regardless of wired or wireless.

  The server device 10 is a computer that provides a Web search service to the client terminal 30. For example, the server device 10 provides a keyword presentation service that presents a narrowing-down keyword that narrows down the number of search results by combining with a keyword that has been designated via the client terminal 30 together with the Web search service. As one aspect, it may be implemented as a Web server that executes processing related to the keyword presentation service, or may be implemented as a cloud that provides the keyword presentation service by outsourcing. As another aspect, it can also be implemented by preinstalling or installing a keyword presentation program provided as package software or online software on a desired computer.

  The client terminal 30 is a computer that receives the keyword presentation service. As an aspect of the client terminal 30, a mobile terminal such as a smartphone, a mobile phone, a PHS (Personal Handyphone System), or a PDA (Personal Digital Assistants) can be employed in addition to a fixed terminal such as a personal computer.

  In the following description, the server device 10 will be described assuming an example in which the server device 10 functions as a portal site that allows the client terminals 30A to 30C to search for Web pages published on the network 5, but the scope of application of the keyword presentation service described above Is not limited to the case where the server device 10 functions as a portal site. For example, when searching for files in a site managed by a specific service provider using, for example, XML (Extensible Markup Language) or HTML (HyperText Markup Language) files described in a markup language using the above keywords The keyword presentation service can be similarly applied. In addition, the above keyword presentation service can be similarly applied to a case where a file stored in a database by an organization such as a company or a group is searched using a keyword for a person concerned in an organization having access authority.

[Configuration of Server Device 10]
Next, the functional configuration of the server device 10 according to the present embodiment will be described. FIG. 2 is a block diagram illustrating a functional configuration of the server apparatus 10 according to the first embodiment. As illustrated in FIG. 2, the server device 10 includes a communication I / F (InterFace) unit 11, a storage unit 13, and a control unit 15. Note that the server device 10 may include various functional units included in a known server device other than the functional units illustrated in FIG. 2, for example, functional units such as various input devices and audio output devices.

  The communication I / F unit 11 is an interface that performs communication control with another device, for example, the client terminal 30. As an aspect of the communication I / F unit 11, a network interface card such as a LAN (Local Area Network) card can be employed. For example, the communication I / F unit 11 receives a keyword used for a search from the client terminal 30, searches for a search using a specified keyword, and further narrows down the search result in combination with the specified keyword. A keyword for narrowing down is transmitted to the client terminal 30.

  The storage unit 13 is a storage device that stores various programs such as an OS (Operating System) executed by the control unit 15 and a keyword presentation program. As one aspect of the storage unit 13, a storage device such as a semiconductor memory element such as a flash memory, a hard disk, or an optical disk can be cited. The storage unit 13 is not limited to the type of storage device described above, and may be a RAM (Random Access Memory) or a ROM (Read Only Memory).

  The storage unit 13 stores an inverted index 13a and a presented keyword route 13b as an example of data used in a program executed by the control unit 15. In addition to the transposed index 13a and the presented keyword route 13b, other electronic data, for example, a cache of a file such as a web page accessed by a web search can also be stored.

  Among these, the transposed index 13a is index data for search. As one aspect of the transposed index 13a, data in which a file ID (IDentifier) including a headword in the text in the file and position information in which the headword appears in the text is associated with each headword as an index is adopted. it can. Examples of the “file” mentioned here include an HTML file, an XML file, a PDF (Portable Document Format) file, and the like. These files only need to contain text as part of them, and may contain images and sounds. By performing a keyword search on the text of files containing images and sounds, It is good also as searching an image and a moving image. In addition, for extracting “headwords”, any index character string extraction method, for example, morphological analysis or N-Gram method can be used.

  By using the transposed index 13a for searching, a file including a keyword designated from the client terminal 30 can be searched without storing the file itself. In the example of FIG. 2, the case where the transposed index 13a is used for the file search is illustrated. However, each file itself is stored in the storage unit 13, and the specified keyword is searched from the text included in each file. It doesn't matter.

  The control unit 15 has an internal memory for storing programs defining various processing procedures and control data, and executes various processes using these. As shown in FIG. 2, the control unit 15 includes a reception unit 15a, a search unit 15b, a creation unit 15c, an extraction unit 15d, a determination unit 15e, and a presentation unit 15f.

  Among these, the reception part 15a is a process part which receives specification of a keyword. As one aspect, when the receiving unit 15a receives an access from the client terminal 30 to a URL (Uniform Resource Locator) that provides the Web search service by the server device 10, the following search screen is displayed on the client terminal 30. Display. Examples of such a search screen include a text box in which a keyword can be input and a pull-down menu in which a keyword can be selected. As described above, the reception unit 15a displays the search screen on the client terminal 30 to receive the specification of the keyword used by the user for the search via the search screen.

  The search unit 15b is a processing unit that searches for a file including a specified keyword with reference to the transposed index 13a. As one aspect, the search unit 15b searches for a file ID associated with a headword corresponding to a keyword that has been designated by the receiving unit 15a among headwords included in the transposed index 13a. Thereafter, the search unit 15b determines whether or not the number of file IDs obtained as a search result, that is, the number of search results is an upper limit value of a predetermined target number, for example, 1000 or less. At this time, if the upper limit of the number of objectives is exceeded, the designated keyword is general, and there is a high possibility that the search results are not narrowed down to the extent that the target information can be browsed. Therefore, the search unit 15b outputs a search result to the creation unit 15c in order to present a narrowing keyword to be combined with the specified keyword. On the other hand, when the number is less than or equal to the upper limit value of the number of objectives, it can be estimated that there is a high possibility that the search result is narrowed down to the extent that the user can browse the target information by the designated keyword. In this case, the search unit 15b outputs the search result to the presentation unit 15f without generating a narrowing keyword.

  The creating unit 15c is a processing unit that creates an appearance map by setting the presence or absence of a specified keyword for each file according to a file search result. As an example of the data structure of such an appearance map, a bit array in which a flag “0” or “1” is set for each file ID can be adopted. For example, the creating unit 15c assigns “1” to the file ID of a file including the specified keyword, while assigning “0” to the file ID of a file not including the specified keyword. In addition to the appearance map of the designated keyword, the creation unit 15c also creates an appearance map of a keyword candidate for narrowing down to be combined with the designated keyword by converting the search result of the keyword candidate for narrowing down into a bit array. .

  The extraction unit 15d is a processing unit that extracts a keyword that achieves a predetermined number of search results for a combination with a specified keyword using an appearance map of the specified keyword and a keyword appearance map combined with the specified keyword. is there. As an aspect, the extraction unit 15d uses a keyword having a co-occurrence relationship with a specified keyword as a keyword candidate for narrowing down in combination with the specified keyword. At this time, the extraction unit 15d adds one keyword candidate for narrowing down to the designated keyword. Then, the extraction unit 15d uses the specified keyword appearance map and the refined keyword candidate appearance map, and the search result of each keyword AND search is equal to or less than the upper limit value of the target number and the lower limit value of the target number. Until this is the case, the process of adding candidate keywords for narrowing down to the specified keyword is repeatedly executed. Thereafter, when the search result of the AND search is equal to or lower than the upper limit value of the target number and equal to or higher than the lower limit value of the target number, the extracting unit 15d further specifies the specified keyword and further the first to Nth narrowing keywords. Are extracted as keyword routes.

  More specifically, the extraction unit 15d sets the specified keyword in the internal memory of the control unit 15 as the first keyword 0 of the keyword route when the appearance map of the specified keyword is created by the creation unit 15c. sign up. At this time, the extraction unit 15d registers the number of search results for the keyword 0 in association with the keyword 0 in the internal memory. Then, the extraction unit 15d lists keywords having a co-occurrence relationship with the designated keyword as candidates for narrowing down keywords to be combined with the designated keyword. Then, the extraction unit 15d selects one new keyword from the keyword candidates for narrowing down. Subsequently, the extraction unit 15d causes the creation unit 15c to create an appearance map of the newly selected keyword. Here, the case where the appearance map is created by the creation unit 15c every time a keyword is newly selected has been illustrated. However, in order to shorten the processing time, an appearance map for each keyword is created and stored in advance. It doesn't matter.

  Thereafter, the extraction unit 15d performs an AND operation of the bit array between the appearance map of the keyword route stored in the internal memory, here, the appearance map of the specified keyword and the appearance map of the newly selected keyword. . That is, the extraction unit 15d performs an AND operation on the values set in the same file ID included in each appearance map. For example, if at least one of the values set for the same file ID is “0”, the value of the file ID is set to “0”. On the other hand, when both values set to the same file ID are “1”, “1” is set to the value of the file ID. After that, the extraction unit 15d counts the number of bits set to “1” in the bit array included in the appearance map obtained by the AND operation. As a result, the number of search results of AND search of each keyword included in the keyword route, that is, keyword 0 and keyword 1, is calculated. Subsequently, the extraction unit 15d adds the newly selected keyword as the keyword 1 of the keyword route to the internal memory. At this time, the extraction unit 15 d associates the keyword 1 with the number of search results of the AND search of the keyword 0 and the keyword 1 and registers them in the internal memory. Then, the extraction unit 15d determines whether or not the number of AND search results of each keyword included in the keyword route is equal to or less than the upper limit of the target number.

  At this time, if the upper limit of the target number is exceeded, there is a high possibility that the search results are not narrowed down to the extent that the target information can be browsed. In this case, the extraction unit 15d further selects a new keyword from the keyword candidates for narrowing down, and then causes the creation unit 15c to create the appearance map, and the appearance map of the keyword route stored in the internal memory. AND operation of bit array is executed between Then, the extraction unit 15d calculates the number of search results of AND search of each keyword included in the keyword route, that is, the keyword 0, the keyword 1, and the keyword 2. Further, the extraction unit 15d adds the newly selected keyword as the keyword 2 of the keyword route to the internal memory. In this case, the extraction unit 15d associates the keyword 2 with the number of search results of the keyword 0, the keyword 1, and the keyword 2 in the AND search and registers them in the internal memory.

  As described above, the extraction unit 15d selects a new keyword from the keyword candidates for narrowing down until the number of search results of the AND search for each keyword included in the keyword route is equal to or less than the upper limit of the target number. Repeat the process to add to the keyword route.

  On the other hand, if it is less than or equal to the upper limit of the target number of cases, it can be estimated that there is a high possibility that the search result is narrowed down to the extent that the user can browse the target information by the designated keyword. In this case, the extraction unit 15d determines whether the number of search results of the AND search for each keyword included in the keyword route is a lower limit value of the target number, for example, 10 or more. In this way, whether or not the target number is equal to or greater than the lower limit of the target number is not more than the upper limit of the target number, but if the number of search results is excessively small, it is important information for the user. This is because there is a high possibility that a file having a will be leaked from the search results.

  And the extraction part 15d erase | eliminates the keyword route memorize | stored in the internal memory, when it is less than the lower limit of the target number of cases. On the other hand, the extraction unit 15d leaves the keyword route stored in the internal memory as it is as a target for determining the validity of the keyword route, which will be described later, when it is equal to or greater than the lower limit of the target number of cases. Thereafter, the extracting unit 15d continues until the search for combining the keyword candidates for narrowing down with the specified keyword is completed for all the combinations, that is, between each keyword constituting the already searched keyword route and the discarded keyword route. Until there are no combinations that do not overlap, the search for combining candidate keywords with the specified keyword is repeatedly executed. At this time, it is clear that the designated keyword is the first keyword 0 of the keyword route. For this reason, the extraction unit 15d starts the iterative process from the stage of selecting the keyword corresponding to the keyword 1 of the keyword route from the keyword candidates for narrowing down.

  FIG. 3 is a diagram illustrating an example of an appearance map of a designated keyword. 4 and 5 are diagrams illustrating an example of an AND operation. FIG. 3 shows an appearance map when “protein” is designated as a keyword. FIG. 4 shows an AND operation of a bit array executed between the appearance map of keyword 0 “protein” and the appearance map of keyword 1 “metabolic”. Furthermore, FIG. 5 illustrates an AND operation of bit arrays executed between the appearance map of keyword 0 “protein” and keyword 1 “metabolic” and the appearance map of keyword 2 “ec”.

  The appearance map of the keyword “protein” shown in FIG. 3 includes the keyword “protein” in each of the file IDs “1”, “3”, “6”, and “7”, and other file ID files. Means that the keyword “protein” is not included. Since the keyword “protein” is a keyword that has been designated from the client terminal 30, it is registered in the internal memory as keyword 0 of the keyword route. If the candidate keyword “metabolic” for selection is selected as keyword 1, an AND operation of the bit array is executed between the appearance map of keyword 0 “protein” and the appearance map of keyword 1 “metabolic”. Is done. As shown in FIG. 4, the values of the file IDs “1”, “3”, and “6” in which the bit “1” is set to the same file ID value between the appearance maps are “1” by AND operation. The other file ID values are calculated as “0”. Here, in the case of only the designated keyword “protein”, the bit of the file ID “7” was “1”, but the bit of the file ID “7” is changed by AND operation with the keyword “metabolic”. It is “0”. This means that the addition of the keyword “metabolic” has excluded the file with the file ID “7” and the narrowing has advanced.

  Furthermore, when the keyword candidate “ec” for narrowing down is selected as keyword 2, a bit is placed between the appearance map of keyword 0 “protein” and keyword 1 “metabolic” and the appearance map of keyword 2 “ec”. An array AND operation is performed. As shown in FIG. 5, the values of the file IDs “3” and “6” in which the bit “1” is set to the same file ID value between the appearance maps are calculated as “1” by the AND operation. On the other hand, the value of the other file ID is calculated as “0”. Here, in the case of the AND operation of the keyword 1 “protein” and the keyword 2 “metabolic”, the bit of the file ID “1” was “1”, but the file ID “1” is obtained by the AND operation with the keyword “ec”. The bit “1” is “0”. This means that the addition of the keyword “ec” excludes the file with the file ID “1” and further narrows down.

  FIG. 6 is a diagram illustrating an example of a keyword route. The first record shown in FIG. 6 means that the number of search results has been narrowed down to 50000 by the keyword 0 “protein” that is the designated keyword. Furthermore, it means that the number of search results is narrowed down to 15000 by keyword 1 “metabolic”. Furthermore, the keyword 2 “ec” means that the number of search results is narrowed down to 900, which is equal to or less than the upper limit “1000” of the target number and is equal to or higher than the lower limit “10” of the target number. The example of the second record shown in FIG. 6 indicates that the number of search results has been narrowed down to 50000 by the keyword 0 “protein” that is the designated keyword. Furthermore, it indicates that the number of search results has been narrowed down to 27,300 by keyword 1 “substrate”. Furthermore, it indicates that the number of search results has been narrowed down to 12,500 by keyword 2 “abc”. Further, the keyword 3 “transport” indicates that the number of search results is narrowed down to 1000 which is equal to or less than the upper limit value “1000” of the target number and is equal to or higher than the lower limit value “10” of the target number.

  Note that here, when the target route count is less than the lower limit, the keyword route stored in the internal memory is discarded, but a rare query that allows the user to browse the target information pinpointed is input. It is possible to leave a keyword route that only determines whether the number of search results is equal to or greater than the upper limit of the target number, and leaves the number of search results less than the lower limit of the target number. .

Returning to the description of FIG. 2, the determination unit 15 e is a processing unit that determines the validity of combining the keyword extracted by the extraction unit 15 d with the specified keyword. As one aspect, the determination unit 15e sets a narrowing rate by which search results are narrowed every time a keyword is added according to the number of keywords combined with a specified keyword. For example, “Z” is the number of search results for the specified keyword, “Q” is the target number, and “α” is the narrowing-down rate at which the search result is narrowed every time one keyword is added. It is assumed that the narrowing rate “α” is constant. At this time, when the keyword combined with the keyword 0 is only the keyword 1, the number of narrowing down narrowed down by the keyword 1 is “Z × α”. In addition, when the keywords combined with the keyword 0 are the keyword 1 and the keyword 2, the number of narrowing down narrowed down by the keyword 1 and the keyword 2 is “Z × α × α”. When the number of keywords to be combined with the keyword 0 is N from the keyword 1 to the keyword N, the number of narrowed down searches by the keyword 1 to the keyword N is “Z × α ^N ”. Therefore, in order to narrow down to Q items by N AND searches, the following equation (1) is established.

Q = Z × α ^N (1)

  Here, an ideal sample value of the number of search results “Z” and the target number “Q” of the specified keyword is set for each number of narrowing keywords to be combined with the specified keyword. As the ideal sample value, an experimental value can be set, or an arbitrary value can be set at the client terminal 30. After that, the determination unit 15e substitutes Z and Q, which are ideal sample values, together with N into the derivation formula of the above formula (1). Thereby, it is possible to derive a narrowing-down rate α by which the search result is narrowed down every time a keyword is added according to the number of keywords combined with the designated keyword. FIG. 7 is a diagram illustrating an example of the relationship between the number of search results for a specified keyword and the number of keywords combined with the specified keyword. In FIG. 7, the number of search results when the number Z of search results by the specified keyword is “10000” and the target number Q is “1000” is shown for each number of keywords combined with the specified keyword. . The vertical axis of the graph shown in FIG. 7 indicates the number of search results, and the horizontal axis indicates the number of keywords combined with the specified keyword. As shown in FIG. 7, when the number of keywords combined with the designated keyword is 1, it is indicated that the narrowing-down rate is preferably about 0.01. Further, when the number of keywords is 2, it is indicated that the narrowing-down rate is preferably about 0.1. Furthermore, when the number of keywords is 3, it indicates that the narrowing-down rate is preferably about 0.215. Furthermore, when the number of keywords is 4, it indicates that the narrowing-down rate is preferably about 0.316. Furthermore, when the number of keywords is 5, it is indicated that the narrowing-down rate is preferably about 0.398. Furthermore, when the number of keywords is 6, it is indicated that the narrowing-down rate is preferably about 0.464. Furthermore, when the number of keywords is 7, it is indicated that the narrowing-down rate is preferably about 0.517.

Under such a configuration in which the narrowing-down rate α is derived, the determination unit 15e compares the number of search results when the keywords up to that order are ANDed for each order from the keyword 1 to the keyword N. to calculate the number W _n narrowed down as a reference. For example, the determination unit 15e acquires the number of keywords combined with a specified keyword. Then, for each order from the keyword 1 to the keyword N included in the keyword route, the determination unit 15e performs the number of times corresponding to the order, that is, the narrowing rate corresponding to the number of keywords combined with the designated keyword over the same number of times as the order. Multiply α by the number of search results for the specified keyword. As a result, the number of narrowed cases W _n in the order is calculated. On top of that, the determination unit 15e, range search volume from the number W _n of predetermined narrow down results when a keyword from the keyword 1 included in the keyword root to the sequence for each sequence to the keyword N is AND operation For example, it is determined whether or not β is within a range of ± 20%. In this case, the search in case the number of results is not in the number W _n ± β narrowing is, narrowing the number W _n whether ± not in the β keyword N is not functioning effectively to narrow down a general word, or keyword It can be seen that N may cause spillage. In this case, it is estimated that the keyword route including the keyword N is not valid. On the other hand, if there is no keyword whose number of search results deviates from the narrowed down number W _n ± β, the use of keywords 1 to N can suppress the loss of keywords, and each keyword 1 to N can function effectively for narrowing down. It can be estimated that the characteristics are high. In this case, it can be estimated that the keyword route including the keyword N is valid. The determination unit 15 e registers the keyword 0 and the keywords 1 to N in the storage unit 13 as the presented keyword route 13 b that is to be presented to the client terminal 30.

  The presentation unit 15 f is a processing unit that performs control of information presentation to the client terminal 30. As an aspect, the presentation unit 15f outputs a list of files including the specified keyword to the client terminal 30 as the search result when the number of search results by the specified keyword is equal to or less than the upper limit of the target number. At this time, the presentation unit 15f can also output a list of files presented as a search result including a link such as a title and URL and an advertisement. As another aspect, the presentation unit 15f outputs a list of files including the specified keyword as a search result, and specifies a keyword included in the presented keyword route 13b registered in the storage unit 13 by the determination unit 15e. Is output to the client terminal 30. Thereafter, when receiving the selection of the next keyword, the presentation unit 15f determines whether or not there is a subsequent keyword in the keyword route. Then, the presentation unit 15 f outputs a search result obtained by performing an AND search on the next keyword previously selected until there is no subsequent keyword, and further outputs the next keyword to the client terminal 30. Thereafter, when there are no more continued keywords, the presentation unit 15 f outputs to the client terminal 30 a search result obtained by performing an AND search on the last keyword.

  8 and 9 are diagrams illustrating examples of screen transitions displayed on the client terminal 30. FIG. Among these, FIG. 8 shows a screen transition in the case where keyword input assistance according to the above-described prior art is performed. Further, FIG. 9 illustrates a keyword route presented by the server device 10 according to the first embodiment. As shown in FIGS. 8 and 9, at the stage of accepting the keyword “protein” accepted from the client terminal 30, the number of search results is 50000 in both the conventional technology and the present embodiment. It has become.

  At this time, in the example of FIG. 8, the keyword “kinase”, which is assisted with input following the keyword “protein”, is selected. As described above, even when an AND search is further performed using the keyword “kinase”, the number has been reduced from 50,000 to 40,000, but only 10,000 are narrowed down. Further, in the example of FIG. 8, the keyword “cancer” that is input-assistant following the keyword “kinase” is selected. As a result, the number of search results has been increased from 40000 to 30000 by performing an AND search with the keyword “cancer”, but only 10,000 results have been narrowed down. Thus, in the conventional technology, even if an AND search is performed with the three keywords “protein”, “kinase”, and “cancer”, a huge search result of 30000 is still displayed, and a predetermined narrowing accuracy is obtained. It may not be possible.

  On the other hand, in the example of FIG. 9, the next keyword “metabolic” presented after the keyword “protein” is selected. In this way, as a result of further AND search with the keyword “metabolic”, the prior art could only narrow down to 40000, but in the present embodiment, it can narrow down from 50000 to 15000. Furthermore, in the example of FIG. 9, the keyword “ec” presented after the keyword “metabolic” is selected. As described above, as a result of performing an AND search with the keyword “ec”, the prior art can only narrow down to 30000, but in the present embodiment, it can narrow down to 900. As described above, in this embodiment, by performing an AND search with the three keywords “protein”, “metabolic”, and “ec”, the search results are narrowed down to the number of cases that the user can view the target information. ing.

  Note that various integrated circuits and electronic circuits can be employed for the control unit 15. In addition, a part of the functional unit included in the control unit 15 may be another integrated circuit or an electronic circuit. For example, an ASIC (Application Specific Integrated Circuit) is an example of the integrated circuit. Examples of the electronic circuit include a central processing unit (CPU) and a micro processing unit (MPU).

[Process flow]
Next, the process flow of the server device 10 according to the present embodiment will be described. Here, (1) the search process executed by the server device 10 will be described, and then (2) the presentation keyword route generation process executed as a subroutine of the search process will be described.

(1) Search Process FIG. 10 is a flowchart illustrating a search process procedure according to the first embodiment. This process is started when a keyword specification is received from the client terminal 30. As illustrated in FIG. 10, when a keyword specification is received from the client terminal 30 by the receiving unit 15a (step S101), the search unit 15b includes a heading corresponding to the specified keyword among the headwords included in the transposed index 13a. The file ID associated with the word is searched (step S102).

  Thereafter, the search unit 15b determines whether or not the number of file IDs obtained as a search result, that is, the number of search results is equal to or less than the upper limit of the target number (step S103). At this time, if the upper limit of the number of objectives is exceeded (No in step S103), the specified keyword is general, and the search result may not be narrowed down to the extent that the target information can be browsed. high. In this case, a process for generating a presented keyword route to be presented to the client terminal 30 as a keyword to be combined with the designated keyword is executed (step S104).

  Then, the presentation unit 15f outputs a list of files including the specified keyword as a search result, and outputs to the client terminal 30 the keyword next to the keyword that has received the specification among the keywords included in the presented keyword route 13b ( Step S105).

  Thereafter, when the presentation unit 15f receives selection of the next keyword (step S106), the presentation unit 15f determines whether or not there is a subsequent keyword in the keyword route (step S107).

  If there is a subsequent keyword (No in step S107), the presentation unit 15f outputs a search result obtained by performing an AND search on the next keyword selected in step S106, and the subsequent keyword is transmitted to the client. It outputs to the terminal 30 (step S105), and further accepts the selection of the next keyword output as a subsequent keyword in step S105 (step S106).

  On the other hand, when there is no subsequent keyword (step S107 Yes), the presentation unit 15f outputs the search result obtained by performing an AND search on the last keyword to the client terminal 30 (step S108), and ends the process.

  If the number is less than or equal to the upper limit of the number of objectives (Yes at Step S103), it can be estimated that there is a high possibility that the search result is narrowed down to the extent that the user can browse the target information with the designated keyword. In this case, the search unit 15b outputs the search result to the presentation unit 15f without generating a narrowing keyword (step S108), and ends the process.

(2) Presentation Keyword Route Generation Processing FIGS. 11 and 12 are flowcharts illustrating the procedure of the presentation keyword route generation processing according to the first embodiment. This process is started when it is determined No (No) in the determination of step S103 illustrated in FIG.

  As illustrated in FIG. 11, the creation unit 15c performs initialization to empty the keyword route stored in the internal memory (step S201). Subsequently, the creating unit 15c creates an appearance map by setting the presence / absence of the appearance of the designated keyword for each file according to the search result by the designated keyword (step S202).

  Then, the extraction unit 15d registers the designated keyword in the internal memory as the first keyword 0 of the keyword route (step S203). At this time, the extraction unit 15d registers the number of search results for the keyword 0 in association with the keyword 0 in the internal memory.

  Then, the extraction unit 15d lists keywords having a co-occurrence relationship with the designated keyword as candidates for narrowing down keywords to be combined with the designated keyword (step S204). After that, the extraction unit 15d selects one new keyword from the keyword candidates for narrowing down (step S205). Subsequently, the extraction unit 15d causes the creation unit 15c to create an appearance map of the keyword selected in Step S205 (Step S206).

  Thereafter, the extraction unit 15d performs an AND operation on the bit array between the keyword route appearance map stored in the internal memory and the keyword appearance map selected in step S205 (step S207). After that, the extraction unit 15d calculates the number of search results of the AND search for each keyword included in the keyword route (step S208).

  Subsequently, the extraction unit 15d adds the keyword selected in step S205 to the internal memory as the keyword 1 of the keyword route (step S209). At this time, the extraction unit 15 d associates the keyword 1 with the number of search results of the AND search of the keyword 0 and the keyword 1 and registers them in the internal memory.

  Then, the extraction unit 15d determines whether the number of search results of the AND search for each keyword included in the keyword route is equal to or less than the upper limit of the target number (step S210). At this time, if the upper limit of the target number is exceeded (No in step S210), there is a high possibility that the search results are not narrowed down to the extent that the target information can be browsed. In this case, the extraction unit 15d further selects a new keyword from the keyword candidates for narrowing down (step S205), and repeatedly executes the processing from step S206 to step S209.

  On the other hand, if it is equal to or less than the upper limit of the target number of cases (Yes in step S210), it can be estimated that there is a high possibility that the search result is narrowed down to such an extent that the user can browse the target information with the designated keyword. In this case, the extraction unit 15d determines whether the number of search results of the AND search for each keyword included in the keyword route is equal to or greater than the lower limit value of the target number (step S211).

  If it is less than the lower limit of the target number (No at Step S211), the extraction unit 15d deletes the keyword route stored in the internal memory (Step S212). On the other hand, if it is equal to or greater than the lower limit value of the target number (Yes in step S211), the extraction unit 15d leaves the keyword route stored in the internal memory as a target for determining the validity of the keyword route, and proceeds to step S213. To do.

  Thereafter, the extraction unit 15d repeatedly executes the processing from step S205 to step S212 until the search for combining the candidate keywords for narrowing down with the specified keyword is completed for all combinations (No in step S213).

  When the search for combining candidate keywords for narrowing down to the specified keyword is completed for all combinations (step S213 Yes), the determination unit 15e selects one keyword route registered in the internal memory (step S214).

Subsequently, the determination unit 15e acquires the number of keywords combined with the designated keyword (step S215). Then, for each order from keyword 1 to keyword N, the determination unit 15e multiplies the number of search results by the designated keyword by the narrowing rate α corresponding to the number of keywords combined with the designated keyword over the number of times corresponding to the order. Thus, the number of narrowed cases W _n in each order is calculated (step S216).

Thereafter, the determination unit 15e performs initialization to set the keyword counter n of the internal memory to the initial value “1” (step S217). Not on its determination unit 15e, or keywords to the order corresponding to the keyword counter n is in the range of the number W _n narrowing the number of search results when the AND operation a predetermined range, for example, of ± beta Is determined (step S218).

At this time, if the number of search results is within the narrowed-down number W _n ± β (step S218 Yes), it is determined whether or not the value of the keyword counter n has become N (step S219). If the value of the keyword counter n is less than N (No in step S219), the keyword counter n is incremented (step S220), and the process of step S218 is executed again. If the number of search results is not within the narrowed-down number W _n ± β (No in step S218), the process from step S219 to step S221 is skipped, and the process proceeds to step S222.

  On the other hand, when the value of the keyword counter n is N (Yes in step S219), the determination unit 15e stores the keyword route selected in step S214 as the presented keyword route 13b to be presented to the client terminal 30. 13 (step S221).

  Thereafter, the process from step S214 to step S221 is repeatedly executed until the validity determination is completed for all keyword routes registered in the internal memory (No in step S222). Then, when the validity determination for all the keyword routes registered in the internal memory is completed (Yes in step S222), the process is terminated.

[Effect of Example 1]
As described above, when the server device 10 according to the present embodiment presents a keyword to be combined with an input keyword, the number of search results satisfies the target number and the combination is valid. Present keywords. Therefore, according to the server device 10 according to the present embodiment, it is possible to improve the search result narrowing accuracy.

  Although the embodiments related to the disclosed apparatus have been described above, the present invention may be implemented in various different forms other than the above-described embodiments. Therefore, another embodiment included in the present invention will be described below.

[Desktop Search]
In the first embodiment, the case where the web search service is provided has been exemplified. However, for example, the above-described keyword presentation service can be similarly applied to a desktop search for searching for a file stored on a computer. it can. As described above, when desktop search is performed, for example, a computer having each function unit in the control unit 15 can be operated in a stand-alone manner instead of the client server system. In this case, for example, file metadata may be stored in the storage unit 13 instead of the file, and a specified keyword may be searched from text included in the metadata of each file. In addition, a word index included in metadata of each of a plurality of files is stored in the storage unit 13 as a metadata transposition index, and a file including a specified keyword is searched using the meta data transposition index. It doesn't matter. Thus, for example, when a part of a character string constituting a file name is specified as a keyword, a character string for further narrowing down a target document file, an image file, or an audio file can be presented.

[Distribution and integration]
In addition, each component of each illustrated apparatus does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. Can be integrated and configured. For example, the reception unit 15a, the search unit 15b, the creation unit 15c, the extraction unit 15d, the determination unit 15e, or the presentation unit 15f may be connected as an external device of the server device 10 via a network. In addition, each of the server devices 10 has a reception unit 15a, a search unit 15b, a creation unit 15c, an extraction unit 15d, a determination unit 15e, or a presentation unit 15f. You may make it implement | achieve a function.

[Keyword presentation program]
The various processes described in the above embodiments can be realized by executing a prepared program on a computer such as a personal computer or a workstation. In the following, an example of a computer that executes a keyword presentation program having the same function as that of the above embodiment will be described with reference to FIG.

  FIG. 13 is a schematic diagram illustrating an example of a computer that executes a keyword presentation program according to the first and second embodiments. As illustrated in FIG. 13, the computer 100 includes an operation unit 110 a, a speaker 110 b, a camera 110 c, a display 120, and a communication unit 130. Further, the computer 100 includes a CPU 150, a ROM 160, an HDD 170, and a RAM 180. These units 110 to 180 are connected via a bus 140.

  As shown in FIG. 13, the HDD 170 is a keyword that exhibits the same functions as the reception unit 15a, search unit 15b, creation unit 15c, extraction unit 15d, determination unit 15e, and presentation unit 15f described in the first embodiment. A presentation program 170a is stored in advance. This keyword presentation program 170a is integrated or separated as appropriate, as with each component of the reception unit 15a, search unit 15b, creation unit 15c, extraction unit 15d, determination unit 15e, and presentation unit 15f shown in FIG. May be. In other words, all data stored in the HDD 170 need not always be stored in the HDD 170, and only data necessary for processing may be stored in the HDD 170.

  Then, the CPU 150 reads the keyword presentation program 170 a from the HDD 170 and expands it in the RAM 180. Thus, as shown in FIG. 13, the keyword presentation program 170a functions as a keyword presentation process 180a. The keyword presentation process 180a expands various data read from the HDD 170 in an area allocated to itself on the RAM 180 as appropriate, and executes various processes based on the expanded data. The keyword presentation process 180a is a process executed by the receiving unit 15a, the search unit 15b, the creation unit 15c, the extraction unit 15d, the determination unit 15e, and the presentation unit 15f illustrated in FIG. Includes the processing shown. In addition, each processing unit virtually realized on the CPU 150 does not always require that all processing units operate on the CPU 150, and only a processing unit necessary for the processing needs to be virtually realized.

  Note that the keyword presentation program 170a is not necessarily stored in the HDD 170 or the ROM 160 from the beginning. For example, each program is stored in a “portable physical medium” such as a flexible disk inserted into the computer 100, so-called FD, CD-ROM, DVD disk, magneto-optical disk, or IC card. Then, the computer 100 may acquire and execute each program from these portable physical media. In addition, each program is stored in another computer or server device connected to the computer 100 via a public line, the Internet, a LAN, a WAN, etc., and the computer 100 acquires and executes each program from these. It may be.

DESCRIPTION OF SYMBOLS 1 Search system 5 Network 10 Server apparatus 11 Communication I / F part 13 Storage part 13a Inverted index 13b Presented keyword route 15 Control part 15a Reception part 15b Search part 15c Creation part 15d Extraction part 15e Determination part 15f Presentation part 30A, 30B, 30C Client terminal

Claims (4)

On the computer,
Accept keyword specification,
A first storage unit that stores an index of a character string included in each of a plurality of files or a plurality of files, or an index of a character string included in metadata of a plurality of files or metadata of each of a plurality of files Referring to the second storage unit to store, search for a file that hits the specified keyword,
Create an appearance map by setting the presence or absence of the specified keyword for each file or file metadata according to the search result of the file,
Using the appearance map of the specified keyword and the appearance map of the keyword combined with the specified keyword, a keyword that achieves a predetermined number of search results of the combination with the specified keyword is extracted;
Get the number of keywords combined with the specified keyword,
For each order of keywords to be combined with the designated keyword, the number of search results by the designated keyword is multiplied by the narrowing rate corresponding to the number of keywords to be combined with the designated keyword over the number of times corresponding to the order. Calculate the number of refinements in the order,
Extraction is performed by determining whether or not the number of search results when the keywords up to the order are ANDed for each order of keywords combined with the designated keyword is within a predetermined range from the number of narrowed down cases. The validity of the keyword is combined with the specified keyword,
A keyword presentation program that presents keywords that are determined to be valid and that executes each process. As a process of extracting the keyword,
2. The keyword presenting program according to claim 1, wherein a keyword whose search result of the combination with the designated keyword is equal to or smaller than a first number and equal to or larger than a second number smaller than the first number is extracted. . Computer
Accept keyword specification,
A first storage unit that stores an index of a character string included in each of a plurality of files or a plurality of files, or an index of a character string included in metadata of a plurality of files or metadata of each of a plurality of files Referring to the second storage unit to store, search for a file that hits the specified keyword,
Create an appearance map by setting the presence or absence of the specified keyword for each file or file metadata according to the search result of the file,
Using the appearance map of the specified keyword and the appearance map of the keyword combined with the specified keyword, a keyword that achieves a predetermined number of search results of the combination with the specified keyword is extracted;
Get the number of keywords combined with the specified keyword,
For each order of keywords to be combined with the designated keyword, the number of search results by the designated keyword is multiplied by the narrowing rate corresponding to the number of keywords to be combined with the designated keyword over the number of times corresponding to the order. Calculate the number of refinements in the order,
Extraction is performed by determining whether or not the number of search results when the keywords up to the order are ANDed for each order of keywords combined with the designated keyword is within a predetermined range from the number of narrowed down cases. The validity of the keyword is combined with the specified keyword,
A keyword presenting method for executing each process for presenting a keyword determined to be valid. A first storage unit that stores an index of a character string included in each of a plurality of files or a plurality of files, or an index of a character string included in metadata of a plurality of files or metadata of each of a plurality of files At least one storage unit among the second storage units to be stored;
A reception unit that accepts keyword specification;
A search unit for searching for a file in which the designated keyword is hit with reference to the first storage unit or the second storage unit;
A creation unit that creates an appearance map by setting the presence or absence of the specified keyword for each file or file metadata according to a file search result;
An extraction unit for extracting a keyword that achieves a predetermined number of search results of a combination with the specified keyword, using an appearance map of the specified keyword and an appearance map of a keyword combined with the specified keyword;
An acquisition unit for acquiring the number of keywords combined with the specified keyword;
For each order of keywords to be combined with the designated keyword, the number of search results by the designated keyword is multiplied by the narrowing rate corresponding to the number of keywords to be combined with the designated keyword over the number of times corresponding to the order. A calculation unit for calculating the number of refinements in the order;
Extraction is performed by determining whether or not the number of search results when the keywords up to the order are ANDed for each order of keywords combined with the designated keyword is within a predetermined range from the number of narrowed down cases. A determination unit for determining the validity of the selected keyword to be combined with the specified keyword;
A keyword presenting apparatus comprising: a presenting unit that presents a keyword determined to be valid.

Images