JP2009075662A - Retrieval support device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To support a user to create a more appropriate search expression by presenting a keyword which is suitable for refinining retrieval results to a user when the keyword is not directly related with an input keyword. <P>SOLUTION: When the input of a search word for retrieving an electronic document is received, a related keyword acquisition means 3 acquires the input search word and a related word input as a search word in the past, and a parallel keyword acquisition means 4 acquires a parallel word connected in parallel with the search word in the electronic document retrieved based on the input search word. Also, the related keyword acquisition means 3 acquires the word related to the parallel word of the search word, and the parallel keyword acquisition means 4 acquires the parallel word related to the search word and the parallel word of the word related to the parallel word of the search word, and presents each related word and the parallel words to the user. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a search support apparatus and a search support program that support search of an electronic document.

An electronic document that matches a keyword input by a user is searched from a document server that stores a large number of electronic documents, and the search result is presented to the user.
Here, regarding the electronic document search support, for example, a document that matches an input keyword and a document that does not match are selected from a document database that holds a plurality of documents, and a word that appears in the selected matching document and an input keyword A document search device that selects a word having a high degree of relevance as a co-occurrence word of a keyword and searches the document database again with a new keyword generated by adding the selected co-occurrence word. An invention of a document search device that divides a group into a plurality of sets, finds co-occurrence words for each set, and searches with a new keyword generated by adding the union of these co-occurrence words to the original keyword Has been proposed (see Patent Document 1).
JP 2004-192374 A

  When searching for an electronic document in a highly specialized field such as the medical field, an appropriate search expression (a set of keywords for realizing a narrowed search) is created for the user to obtain a desired search result. It becomes important. However, it is not easy to create an appropriate search formula for those who have insufficient knowledge in the field, such as inexperienced intentions.

  As a technique for solving this problem, a technique using a search expression used in the past has been put into practical use. For example, Google Suggest (http://www.google.co.jp/webhp?complete=1&hl=en) automatically completes search keywords based on the history of past search expressions entered in the search system. It has a function to do. In such a system, when the keyword “diffusion weighted image” is input, past search formula history such as “delayed phase”, “fat”, “high signal”, “axial position”, etc. as candidate keywords that follow. It is possible to automatically display frequently co-occurring keywords and to support the creation of appropriate search expressions.

  However, in the above technique, a keyword input by the user (in the above example, “diffusion weighted image”) and a keyword that frequently co-occurs in the history of the search formula input in the past (in the above example, “delayed phase”). “Fat” “High Signal” “Axial Position”) are only displayed. Therefore, even if a keyword is highly relevant to the input keyword and can be appropriately refined by using it together with the input keyword, it is a keyword that does not co-occur with the input keyword in the history of past search expressions. If there is, it will not be displayed. Further, even in the case of a system that complements keywords based on the appearance frequency of words in an electronic document as in the invention described in Patent Document 1, only a list of keyword candidates is listed, and appropriate keywords are listed. The problem remains that it is difficult to specify the combination.

  The present invention has been made in view of the above-described conventional circumstances. Even if the keyword is not directly related to the input keyword, the keyword is selected if it is a suitable keyword from the viewpoint of narrowing down search results. It is intended to assist the user in creating a more appropriate search expression by presenting it to the user.

  The present invention according to claim 1 is a first receiving means for receiving an input of a search word for searching an electronic document from a user, a storage means for storing an input history of the search word, and the input search First acquisition means for acquiring a related word that was previously input as a search word together with a word, or a parallel word that is connected in parallel to the search word in an electronic document that is searched based on the input search word; Related words previously input as search words together with the parallel words acquired by the first acquisition means, or parallel words connected in parallel to the related words acquired by the first acquisition means in the electronic document to be searched And a presenting means for presenting related words and parallel words obtained by the first and second obtaining means to a user. .

  According to a second aspect of the present invention, in the search support device according to the first aspect, the parallel word connected in parallel to the related word acquired by the second acquisition means in the electronic document to be searched is acquired. 3, and the presenting means further presents the parallel word acquired by the third acquiring means to the user.

  According to a third aspect of the present invention, in the search support apparatus according to the first or second aspect, the designation of the presented word is used to be used for searching an electronic document together with the input search word. Second receiving means for receiving information from a person, and the storage means stores the input search word and the specified word in association with each other as an input history.

  According to a fourth aspect of the present invention, in the search support device according to any one of the first to third aspects, the presenting means includes the related words acquired by the first and second acquisition means, and It is characterized by presenting parallel words in a hierarchical structure.

  The present invention according to claim 5 is a first receiving means for receiving an input of a search word for searching an electronic document from a user, a storage means for storing an input history of the search word, and the input A first acquisition of a related word that has been input as a search word in the past together with the search word that has been input or a parallel word that is connected in parallel to the search word in an electronic document that is searched based on the input search word And a related word previously input as a search word together with the parallel word acquired by the first acquisition means, or a related word acquired by the first acquisition means in the electronic document to be searched The search support program for functioning as a presenting means for presenting the user with the second acquisition means for acquiring the parallel words and the related words and parallel words acquired by the first and second acquisition means.

  According to the search support device according to claim 1, not only a keyword directly co-occurring with an input keyword, such as a related word or a parallel word of an input keyword (search word), but also a parallel word of a related word of an input keyword A keyword that is not directly co-occurring with the input keyword but is suitable for narrowing down the search result, such as a related word of a parallel word of the input keyword or the input keyword, is also presented to the user. Therefore, when creating a search expression for narrowing down the search results, the user can easily create a search expression for obtaining a desired search result because the range of selection of keywords used for narrowing down the search results is widened.

  According to the search support device of the second aspect, it is possible to further expand the range of selection of keywords used for narrowing down the search results.

  According to the search support device of the third aspect, the result of the current search support can be reflected in the search support to be performed later.

  According to the search support device of the fourth aspect, it becomes easy to grasp the relevance of each keyword used for narrowing down the search result.

  According to the search support program of the fifth aspect, the search support device that exhibits the above-described effects can be realized using a computer.

The present invention will be specifically described based on an embodiment exemplified below.
FIG. 1 shows a functional block diagram of a search support apparatus according to this example.
The search support apparatus of this example includes a search target text storage unit 8 that stores a plurality of electronic documents as search target texts, and a search keyword input unit that receives input of keywords (search terms) for searching electronic documents from a user. 1. Search expression history storage means 2 for storing a history of search expressions (keyword sets) input in the past; related keyword acquisition means 3 for acquiring related keywords (related words) of input keywords from past search expression history; Parallel keyword acquisition means 4 for acquiring parallel keywords (parallel words) of input keywords from an electronic document searched based on the input keywords, search keyword structuring means 5 for hierarchically collecting a set of acquired related keywords and parallel keywords, Entering a search expression to present a hierarchically structured keyword set to users and narrow down the search results Search expression input means 6 for accepting from the user, and a search unit 7, to find the electronic document from the search text storage means 8 based on the input search expression.

The related keyword acquisition unit 3 acquires a keyword input in the past together with the input keyword received by the search keyword input unit 1 as a related keyword.
The related keyword acquisition unit 3 also acquires related keywords input in the past together with the parallel keyword of the input keyword acquired by the parallel keyword acquisition unit 4.

The parallel keyword acquisition means 4 is connected to the electronic document searched based on the input keyword with the input keyword and a parallel particle / parallel conjunction (“to” “ka” “ya” “·” “or” etc.) To obtain a noun or compound noun as a parallel keyword. It is also possible to acquire parallel keywords by parsing sentences in an electronic document to identify parallel relations (for example, [Hiroshi Masuichi, Tomoko Okuma, “A practical Japanese analysis system based on Lexical Functional Grammar” ”, Natural Language Processing, 2003, Vol. 10, No. 2, p. 79-109].
The parallel keyword acquisition unit 4 acquires parallel keywords connected in parallel to the related keyword of the input keyword acquired by the related keyword acquisition unit 3 and parallel keywords connected in parallel to the related keyword of the parallel keyword of the input keyword. Also do.

  That is, the related keyword acquisition unit 3 and the parallel keyword acquisition unit 4 acquire the related keyword 12 of the input keyword 11 and the parallel keyword 13 of the related keyword 12 as shown in FIG. 14 and the related keyword 15 of the parallel keyword 14 and the parallel keyword 16 of the related keyword 15 are acquired.

The search support processing by the search support device of this example will be described with reference to the processing flowchart shown in FIG.
When the keyword input from the user is received by the search keyword input unit 1 (step S1), the related keyword acquisition unit 3 acquires a set WR of the related keywords (Wr1, Wr2,..., Wrn) of the input keyword. (Step S2), the parallel keyword acquisition unit 4 acquires a set WP of parallel keywords (Wp1, Wp2,..., Wpm) of the input keyword (Step S3).

Moreover, the parallel keyword acquisition means 4 performs the process A which extracts the set of the parallel keyword about each element of the related keyword set WR of an input keyword (step S4).
Moreover, the related keyword acquisition means 3 acquires the set WPP of the related keywords (WP1, WP2,... WPm) for each element of the parallel keyword set WP of the input keyword (step S5), and the parallel keyword acquisition means. 4 performs a process B for acquiring the set of parallel keywords for each element of the related keyword set WPP of the parallel keywords of the input keyword (step S6).

  After that, the search keyword structuring means 5 obtains a product set for every combination of the two sets for all keyword sets obtained in the process A and the process B, extracts different product sets, and inputs a search expression The means 6 presents the input keyword together with the sum of the parallel keyword set WP of the input keyword and accepts an input of a search expression for narrowing down the search result from the user (step S7).

Hereinafter, the search support process will be described with a specific example.
For example, when a “diffusion weighted image” is accepted as an input keyword, and a search expression history as shown in FIG. 4 is stored in the search expression history storage means 2, the related keyword acquisition means 3 causes the input keyword “ “Delayed phase”, “high signal”, and “axial position” are acquired as related keywords of “diffusion weighted image”.
If the electronic document of FIG. 5 is searched based on the input keyword “diffusion-weighted image”, the parallel keyword acquisition unit 4 performs “T1-weighted image” “T2-weighted image” as parallel keywords of the input keyword “diffusion-weighted image”. Image "and" Fat suppression image "are acquired.

  Similarly, for the related keyword “delayed phase” of the input keyword “diffusion weighted image”, for example, “early phase”, “equilibrium phase”, “portal vein phase”, “arterial phase”, “venous phase”, and related keyword “axial position” For example, “oblique position”, “crossing”, “sagittal section”, and “crown section” are acquired as parallel keywords. Furthermore, for the parallel keyword “T1 weighted image” of the input keyword “diffusion weighted image”, for example, “early phase”, “calcification”, “low signal”, and “coronary break” are acquired as related keywords by the related keyword acquisition unit 3. , Those parallel keywords are also obtained. The same applies to the parallel keywords “T2 weighted image” and “fat suppression image” other than the input keyword “diffusion weighted image”.

  In other words, the related keywords “delayed phase”, “high signal”, “axial position” of the input keyword “diffusion weighted image”, and parallel keywords “T1 weighted image”, “T2 weighted image”, “fat suppression image”, and the like. In addition to the keywords that co-occur with the keyword, keywords that co-occur with these keywords, that is, keywords that co-occur with the input keyword are also acquired.

Thereafter, the search keyword structuring means 5 extracts the input keyword and the parallel keyword set X of the input keyword, and compares the combinations of the sets with respect to the parallel keyword set of the related keywords of each keyword in the set M. Then, a set Y whose set intersection is not an empty set is extracted.
That is, in the case of the above example, {Diffusion weighted image], “T1 weighted image”, “T2 weighted image”, “Fat suppression image”} are extracted as set X, and {Y ’ "Low signal", "Equal signal", "Medium signal", "No signal"} (Set Y1), {"Axial position", "Slanting position", "Transverse", "Sagittal section", "Coronary section" } (Set Y2), {"delayed phase", "early phase", "equilibrium phase", "portal vein phase", "arterial phase", "venous phase"} (set Y3).

The keyword sets X and Y (Y1 to Y3) of the input candidates obtained in this way are presented to the user by the search formula input means 6 on a search formula input screen as shown in FIG.
The search expression input screen shown in FIG. 6 includes an input keyword display unit 21, a keyword specification unit 22 that displays a set of input candidate keywords in a hierarchical structure, and accepts specification of keywords used for search from a user. A search button 23 for receiving a search instruction based on a keyword-based search formula from a user is provided.

  According to the drawing, each element of the keyword set Y1, Y2, Y3 is displayed according to the set subordinate to the input keyword “diffusion emphasized image” which is one element of the keyword set X, and other elements of the keyword set X are “ For each of the T1-weighted image, the T2-weighted image, and the fat suppression image, the elements of the keyword sets Y1, Y2, and Y3 are displayed for each set depending on the keywords.

When a keyword to be used for search is designated (in this example, a radio button provided on the left side of each keyword is checked) and a search instruction is given, the search means 8 is an electronic document corresponding to a search expression based on the specified keyword. Is retrieved from the search target text storage means.
In this example, a proximity search is performed for each keyword in the keyword set X ({“Diffusion weighted image”, “T1 weighted image”, “T2 weighted image”, “Fat suppression image”}), and the obtained results are crossed. Is the final result. For example, the keyword “Diffusion-weighted image” of the set X and the “delayed phase” of the set Y related to the keyword are designated, and the keyword “T1 weighted image” of the set X and the “delayed-phase” of the set Y related to the keyword are specified. "Diffusion-weighted image" and "delayed phase" appear within N words, and "T1-weighted image" and "crossing" appear within N words. (N is a preset positive integer threshold value). In addition, you may perform a normal full text search, without performing such a proximity search.
The searched electronic document is presented to the user, and accordingly, each keyword used for the search is associated with the input keyword and stored in the search expression history storage unit 2 as a search expression history.

In addition, although the case where the input of one keyword was received by the search keyword input means 1 was demonstrated, when the input of a some keyword is received, a process should just be performed about each of the said received keyword.
In this example, the input candidate keyword set is displayed in a hierarchical structure as described above, but each element of the keyword set may be displayed in a simple list form.
Also, in this example, the keyword set of input candidates is presented to the user by displaying the above search expression input screen on the display screen provided on the terminal operated by the user. May be printed out on a medium such as paper and presented to the user.

FIG. 7 shows the main hardware configuration of the search support apparatus of this example.
That is, the search support apparatus of this example includes a CPU that performs various arithmetic processes, a RAM that is a work area of the CPU, a ROM that stores basic control programs, a search support program for realizing each function according to the present invention, and the like. Communicating with other devices such as an HDD that stores information, a liquid crystal display that displays and outputs information to the user, an input / output I / F that is an interface with devices such as a mouse and a keyboard that accept input of information from the user It is comprised with the computer which has hardware resources, such as communication I / F which is an interface which performs.

  Then, the retrieval support program according to the present invention is read from the HDD, expanded in the RAM, and executed by the CPU, so that the storage means (retrieval history storage means 2) according to the present invention, the first reception means (retrieval means (search) Keyword input means 1), second reception means (search expression input means 6), first acquisition means (related keyword acquisition means 3 and parallel keyword acquisition means 4), second acquisition means (related keyword acquisition means 3 and Parallel keyword acquisition means 4), presentation means (search expression input means 6) and the like are realized in a computer of the search support apparatus.

The program according to the present invention is provided to the practitioner of the present invention, for example, in a format for distributing an external storage medium such as a CD-ROM storing the program or a format for distributing via a network.
In addition, each functional unit of the search support apparatus according to the present invention is not limited to a mode realized by the software configuration as in this example, and may be configured by a dedicated hardware module.
In addition, each functional unit of the search support apparatus according to the present invention is not limited to an aspect provided in one computer as in this example, and may be provided distributed in a plurality of computers.

It is a functional block diagram of the search assistance apparatus which concerns on one Embodiment of this invention. It is a figure which shows the hierarchically structured keyword set which concerns on one Embodiment of this invention. It is a processing flowchart of the search assistance process which concerns on one Embodiment of this invention. It is a figure which illustrates the search type | formula log | history which concerns on one Embodiment of this invention. It is a figure which illustrates the content of the electronic document which concerns on one Embodiment of this invention. It is a figure which shows the example of a display of the search type | formula input screen which concerns on one Embodiment of this invention. It is a hardware block diagram of the search assistance apparatus which concerns on one Embodiment of this invention.

Explanation of symbols

1: Search keyword input means,
2: Search expression history storage means,
3: Related keyword acquisition means,
4: Parallel keyword acquisition means,
5: Search keyword structuring means,
6: Search expression input means,
7: Search means,
8: Search target text storage means

Claims (5)

First receiving means for receiving an input of a search term for searching an electronic document from a user;
Storage means for storing a search term input history;
A related word input as a search word in the past together with the input search word or a parallel word connected in parallel to the search word in an electronic document searched based on the input search word Acquisition means of
Related words previously input as search words together with the parallel words acquired by the first acquisition means, or parallel words connected in parallel to the related words acquired by the first acquisition means in the electronic document to be searched Second acquisition means for acquiring
Presenting means for presenting related words and parallel words acquired by the first and second acquiring means to a user;
A search support apparatus comprising: A third acquisition unit that acquires a parallel word connected in parallel to the related word acquired by the second acquisition unit in the electronic document to be searched;
The search support apparatus according to claim 1, wherein the presenting unit further presents the parallel word acquired by the third acquiring unit to the user. A second receiving means for receiving a designation of the presented word from a user for use in searching for an electronic document together with the input search word;
The search support device according to claim 1, wherein the storage unit stores the input search word and the designated word as an input history in association with each other.   The said presentation means presents the related word and parallel word which were acquired by the said 1st and 2nd acquisition means in the aspect which carried out the hierarchical structure, The any one of Claim 1 thru | or 3 characterized by the above-mentioned. Search support device. Computer
First receiving means for receiving an input of a search term for searching an electronic document from a user;
Storage means for storing a search term input history;
A related word input as a search word in the past together with the input search word or a parallel word connected in parallel to the search word in an electronic document searched based on the input search word Acquisition means of
Related words previously input as search words together with the parallel words acquired by the first acquisition means, or parallel words connected in parallel to the related words acquired by the first acquisition means in the electronic document to be searched Second acquisition means for acquiring
The search assistance program for functioning as a presentation means to show a user the related word and parallel word which were acquired by the said 1st and 2nd acquisition means.

Images