JP2000331020A - Method and device for information reference and storage medium with information reference program stored - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a means which efficiently accesses objective information at the time of referring to this objective information, which a user wants, out of numerous information. SOLUTION: A user 1 finds a search object for information which the user wants to refer to and acquires a reference history of information, which can be candidates of the response to this search object from numerous information sources and acquires information, which can be the response to the object from this reference history. Next, the information which can be the response is divided and divided information are stored in a knowledge DB 4 together with the search object in the order of appearance. If a new reference object exists, the search object closest to it is retrieved from search objects stored in the knowledge DB 4, and information meeting the retrieved object are extracted and are displayed in the order.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer-based information reference method, and more particularly, to a user searching for information for a certain purpose by using a past problem solving history.
The present invention relates to a system for efficiently referencing information.

[0002]

2. Description of the Related Art At present, the World Wide Web (hereinafter, World Wi
de Web is described as WWW), the information is exploding. Such information can be regarded as an aggregate of knowledge, and various methods have been devised for efficiently using this information. Hereinafter, three typical conventional systems for utilizing this information will be described.

[0003] First, there is a system that classifies information from a subjective point of view so that it can be referred to efficiently. This is called a directory search service. Taking sports information as an example, various types of sports are classified under the concept of sports. Soccer, baseball,
Since information such as rugby is consistent with the general common sense that it exists in the hierarchy under sports information, it is considered that users can access such information without being confused.

[0004] The point of this service is that various kinds of information are put together into appropriate concepts and arranged hierarchically. As a representative example, NTT Direct
ory (http://navi.ntt.co.jp), Yahoo (htt
p: //www.yahoo.co.jp).

Next, there is a system called a robot type search service. It is said that hundreds of millions of information currently exist on the WWW. It is said that realization of such information by the directory-type search service is considerably difficult because the classification method cannot be automated. Therefore, in the robot type search service, WW
In order to search for miscellaneous information on the W, information on the WWW is collected as much as possible in advance by automatically following hyperlinks of HTML (Hyper Text Markup Language) which is a description format of the WWW information. By making the collected information searchable using a full-text search function, W
The retrieval of information on the WW is virtually realized. TITA
N (http://titan.mcnet.ne.jp/), Good (http: // ww
w.goo.ne.jp) is a typical example.

Finally, a description will be given of a collaborative filtering service which is a fundamentally different service from the above two services. Both the directory-based search service and the robot-based search service are basically performed based on one's own sensitivity when searching for information. On the other hand, the collaborative filtering service realizes more efficient information access by utilizing the knowledge of other users. In this service, for example, when a user is searching for a music CD, it is possible to recommend information similar to that CD. This is to allow other users to evaluate whether they like or dislike a certain CD, to statistically calculate the tendency of preference for each user, and to newly recommend a CD of interest to users with similar preferences. This makes it possible to find bargain CDs. firefly (http: //www.firefly.
com /) is a typical example.

[0007]

Problems to be Solved by the Invention (1) Directory type search service, (2) Robot type search service, (3)
The problems with each of the collaborative filtering services are described below.

(1) In the directory type search service,
Due to the characteristics of a system that classifies information subjectively, there is a problem that the accuracy of information is high but the amount of information is small. Currently, only about 100,000 items of information are collected. In this case, the ratio of all information on the WWW reaches only a few percent,
In some cases, information is missed or the information that is really needed cannot be reached.

[0009] (2) In the robot type search service, software called Internet robots is used for H search.
Information is collected by recursively following hyperlinks of TML. In this case, it is possible to collect a considerably large amount of information as compared with the directory type search service of (1). It is now possible to search. In this system, the user expresses the search purpose by a search term and outputs the search terms in order from the one closest to the search term. However, if the search term is not carefully selected, a huge number of results are returned. As a result, a new problem of searching for necessary information from the information has arisen.

(3) In the collaborative filtering service,
An effective information recommendation service using the knowledge of another user can be realized. However, information recommendation is performed independently of information recommended by other users, and the relationship between a plurality of pieces of recommended information is sparse. In other words, there is a high possibility that highly accurate information is recommended, but there is no meaning in the order of appearance, and there is a problem that it is not possible to know when information that is really desired is recommended.

An object of the present invention is to solve the above-mentioned problems and to provide means for enabling efficient access to information.

[0012]

SUMMARY OF THE INVENTION The present invention provides a method for efficiently referring to information intended by a user from among a vast amount of information using information possessed by other users. In the information reference method, the search purpose for the information that the user wants to refer to, that is, the purpose of searching for the information, is found, and the search purpose for the information becomes a candidate for the answer from a vast amount of information sources. A reference history for the information to be obtained is acquired, the acquired reference history is stored in the reference history storage unit, information that can be an answer to the purpose is obtained from the accumulated reference history, information that can be the answer is divided, and the divided information is divided. Information is stored in the knowledge database together with the search objectives in the order in which they appear, and when there is a new objective to be referenced, the closest search objective among the search objectives stored in the knowledge database Find the target, from the results, extracts information that matches the search object, the extracted information, and key display means displays along its sequence.

The operation of the present invention is as follows. The present invention realizes efficient access to information by utilizing a series of flows of solving a problem of another user. That is,
If a similar search has been performed in the past for the purpose of information search, the reference history is acquired, and further information is referred to, so that the problem solving process of other users can be directly performed. There is an advantage that it can be used.

A program for realizing the above processing by a computer can be stored in an appropriate storage medium such as a computer-readable portable medium memory, a semiconductor memory, and a hard disk.

[0015]

BEST MODE FOR CARRYING OUT THE INVENTION In the following, for convenience of explanation, WWW
An example of application in a system using the (World Wide Web) will be described. However, the present invention is not limited to the WWW, but can be applied to a fragmentary information and an information space in which a link for linking information exists. It is possible to apply.

FIG. 1 is a system configuration diagram for realizing the present method. Each user (client) 1 refers to the information 5 via the relay server 3 with respect to the information on the WWW 6. The access history at that time is stored in the reference history storage unit 2.
Is accumulated in Further, the knowledge generated there is also accumulated in the knowledge database 4 (hereinafter, the database is referred to as DB
Described). Details of the knowledge DB 4 will be described later.

FIG. 2 is a block diagram showing a main part of the embodiment of the present invention. In the drawing, those having the same reference numerals as those in FIG. 1 correspond to those shown in FIG. 1, and reference numeral 7 denotes a WWW browser for extracting and viewing information 5 on WWW 6.

In the relay server 3 having a search function, the search purpose extracting means 30 is a means for finding the search purpose for the information possessed by the user 1 from the search phrase input by the user 1 from the WWW browser 7 or the like. is there. The reference history obtaining means 31 obtains a reference history for information that can be a candidate for the answer from a huge amount of information sources on the WWW 6 for the found search purpose, and accumulates the reference history in the reference history storage unit 2. is there.

The answer candidate information acquiring means 32 is means for acquiring information that can be an answer to the search purpose from the reference history stored in the reference history storage unit 2. Information division means 3
Reference numeral 3 denotes a unit that divides information that can be an answer based on the expression format of the information. In this example, the information dividing means 33
Has an unnecessary information exclusion unit 331 that rejects unnecessary information that can be presumed to be outside the search purpose from among the divided information.

The knowledge accumulating means 34 is means for accumulating, in the knowledge DB 4, information suitable for the search purpose among the information divided by the information dividing means 33 together with the search purpose in the order of appearance. In this example, the knowledge storage means 34 has an appearance frequency recording means 341 for recording the appearance frequency of the information stored in the knowledge DB 4.

The search purpose search means 35 analyzes the search request from the user 1 by the search purpose extraction means 34 and finds a new search purpose to be referred to, from the search purposes stored in the knowledge DB 4. A means for searching for the closest search purpose. When searching for this search purpose, the search purpose search means 35 searches the one having a high appearance frequency with a weight.

The information extracting means 36 is used for the search purpose searching means 3.
This is a means for extracting information with an order suitable for the search purpose searched in step 5 from the knowledge DB 4 or the reference history storage unit 2 or another information source via the network.
The information display means 37 is means for sending the information with the order extracted by the information abstraction means 37 to the WWW browser 7 of the user 1 and displaying the information in the order.

The information extracting means 36 checks whether or not a part of the information corresponds to a part of the information with another order when extracting the information with the order corresponding to the search purpose. If the information display means 37 finds matching information as a result of this check, it displays the information simultaneously or sequentially as one of the options when displaying the result.

FIG. 3 is a flowchart for implementing the present method. In the following, a detailed description will be given using these. For simplicity, the description will be given based on the information search situation of a certain user on the WWW 6 as an example.

When we come across something we don't understand, we may try to get an answer from WWW6,
Sometimes they seek to get information from other sources. For example, information from newspapers and magazines, information from other users, etc., which can be used to answer questions in the real world, are often found to be more effective. In this case, such information is not taken into account, and only the information on WWW 6 is used. However, the present method can be extended to other cases. There are various kinds of information 5 on the WWW 6, from which the user obtains valid information.

When searching for necessary information from the WWW 6, a search service is often used. Search services used in such a case include a directory search service and a robot search service. FIG. 4 shows a state of information search using these services. For these services, the actual search is performed after converting what the user wants to know into search terms.

In this example, when there is a search request "I want you to teach professional baseball camp information", a search term "professional baseball" is input to a certain search service.
As a result, several pieces of information that satisfy the search term are displayed. The user refers to a number of pieces of information in order to find the most suitable information for his or her search purpose. Until the answer is finally obtained, such operations are repeated, and the information referred to at a certain time may be exact, or the problem may be solved by collecting the referred information in pieces.

In this method, the process of solving a problem as described above is taken into the knowledge DB 4 in advance. First, it is assumed that the user 1 is accessing the information 5 on the WWW 6 via the relay server 3. At some point, if an unfamiliar matter appears for the user 1, the user 1 takes various strategies to solve the problem. When an attempt is made to obtain the answer from information on the WWW 6, search information is often used to search for appropriate information.

As an example, it is assumed that the user wants to know “camping information of professional baseball” as shown in FIG. 4, and the search service is used to solve questions such as “professional baseball” and “camping information of professional baseball”. Search terms for the search. The search service receives the search term and displays it to the user in the order of the most appropriate information. The results are often large, but users will see information that they consider appropriate.

As described above, the purpose of the search can be found by the relay server 3 watching the access to a specific search service and the search term entered therein (step 101 in FIG. 3). To do so, a list of search services that need to be monitored is created in advance, and if a corresponding reference activity is found in the reference history of the relay server 3,
Assuming that the user has performed a search activity, a search term to be entered therein is extracted.

In the above example, "professional baseball" and "professional baseball camp information" correspond to these, and these are extracted by the morphological analysis program only as words that are likely to be keywords, and are regarded as search purposes. For example, professional baseball is broken down into "pro" and "baseball". In order to find the search purpose as described above, the search services are listed in advance in the relay server 3.

For example, the following search service can be considered. 1. Directory type search service NTT Directory (http: //navi.ntt.c
o.jp) Yahoo (http://www.yahoo.co.jp) Robot-type search service TITAN (http://titan.mcnet.ne.jp/) Goo (http://www.goo.ne.jp) ・ ・ ・ ・ In relay server 3, on WWW6 from user From the reference requests for the information 5 described above, a reference request for the information represented by the URL is extracted, and more specific request contents are also extracted.
This extraction can be performed empirically for each search service.

For example, in TITAN, a user's search request is: http://titan.mcnet.ne.jp:30000/titan-ad.html?IPSA_ALL=java+sun\&IPSA_MOD=strict & IPSA_SRH = ej & IPSA_MXN = 100 & IPSA_PAG = japanese, and the actual search phrase is represented by a character string after IPSA_ALL. This character string can be changed to a human-readable state by a simple operation.
In the above example, it is indicated that the search term “java and sun” has been input. The search purpose is generated by extracting only words and phrases that are likely to be keywords from this search word, and here, the words “java” and “sun” correspond.

After inputting a search phrase to the search service as shown in FIG. 4, the user selects information suitable for achieving his or her search purpose from return information of the search service. I will refer to it. That is, after such a search term is input and before the next search term is input, it can be regarded as an information group having a high probability that an answer to the search purpose exists. A reference history following such a search term input is acquired (step 102 in FIG. 3).

As described above, information that can be regarded as an answer is considered to appear not only in a series of information until the next search term is input as described above, but also in various cases. For example, if the reference history is analyzed, it can be concluded that the user arrives at the target information between successive references and references that have been spaced a long time before the reference, and that the information is their own. By thinking, one wall can be considered to exist.

When the reference histories are obtained in this manner, information corresponding to each of the histories is actually obtained (step 103 in FIG. 3). If it exists as the cache information of the relay server 3, it may be obtained from there, and if not, it is actually obtained via the network. Such a selection is realized by leaving it to the cache server.

As described above, if information having a high probability of being an answer corresponding to the user's search purpose can be obtained, each piece of information is divided (step 104 in FIG. 3).
This division is performed by the information expression format. For example, if the information is described in HTML, HR, H? (? =
HTM whose text contents are likely to change, such as 1-6)
Information can be divided using the L tag as a mark.

FIG. 5 shows an example of the HTML source file. In this example, the information is sports information as a whole, but information on baseball, soccer, and rugby is scattered among them, and these can be distinguished by a header called H3. Therefore, when this is used as a key to divide the information into baseball, soccer, and rugby information. Such information division can be performed more accurately if the meaning of the information is considered.

FIG. 6 shows an example of information division. User 1
Refers to the information A, B, and C on the WWW 6 continuously based on the result of a certain search service. Each piece of information is A-1 to A-4, B-
1 to B-6 and C-1 to C-3, and the user 1 refers to B by following the hyperlink indicating B-5 existing in A-2. And B-5
Is selected from C. Which part is referred to can be determined by extracting hyperlink information from the immediately preceding information in the reference history and selecting information including a part that matches the information. The partial reference histories for the individual information thus segmented are stored together with the search purpose as the knowledge DB 4 (step 105 in FIG. 3).

FIG. 7 is a diagram showing an example of the knowledge DB. A reference history and partial reference information are stored for each search purpose.
In this example, for example, when the search purpose is “professional” or “baseball”, a history referred to by a certain user and a part considered to be particularly referred to in the information are stored as partial reference information. In this example, the partial reference information is a serial number indicating the number of the divided information, but the information itself may be stored. How to accumulate may be determined by the efficiency.

It is assumed that a user is searching for information on the WWW 6 to solve a problem in a state where the information is stored in the knowledge DB 4 as described above. Relay server 3
In, all the reference histories using the relay server 3 are accumulated. The method starts processing when the user accesses a search service listed in advance. First, the search purpose is extracted. This performs the same processing as when the knowledge DB 4 was created. That is, from the search requests to the listed search services, only search terms are extracted by using empirical rules, and then only those that can be keywords are extracted. With respect to the search objectives thus obtained, similar search objectives are sequentially extracted from the knowledge DB 4 (step 106 in FIG. 3).

As the extraction method, a conventional search method may be used. For example, a simple keyword matching algorithm can be used. In this extraction, it is necessary to sequentially output the objectives based on the precision of the match. At this time, if a search is performed by assigning weights to those having a high appearance frequency, it is possible to preferentially extract information referred to by many users.

If there is appropriate information (step 107 in FIG. 3), the corresponding partial reference history information is
Extract while maintaining the order (Step 10 in FIG. 3)
8). The obtained information is sequentially presented to the user (step 109 in FIG. 3). The presentation method may be a conventional method, and information is displayed in chronological order. By sequentially referring to the presented information, the user can efficiently refer to information that the previous user has traced to similar problems.

FIG. 8 shows a display example of the result by the information display means 37. For example, assuming that the search term input from user 1 is “about baseball performance”, first,
A search keyword is extracted from the search phrase by morphological analysis. On the other hand, the best match is selected from the search objectives stored in the knowledge DB 4 in advance. Here, it is assumed that it corresponds to the history Z. This history Z
Since the partial reference histories are A-3, B-3, and C-2, the user 1 displays the information in the WWW browser 7 in this order, so that the user's history can be mainly referred to. it can.

When displaying the history, not only the history searched for the search purpose is displayed, but also the following expansion is possible. FIG. 9 shows an example of accumulation in the knowledge DB. Information D extracted from the histories of various users
When the data is stored in B4, the same information may be traced on the way.

For example, the information traced by the user U1 in FIG. 9 shares part of the information traced by the user U2 (A
-2, B-5). For the user U1, the traced information (C-1) is correct, and for the user U2, the user U1 is also correct.
It is considered that the information (C-3) is correct, not the information (C-1) traced by the user. For a user who later refers to these histories, if the search purposes of the histories X and Y are similar, it is highly likely that which one is more appropriate according to the situation. Therefore, when a plurality of pieces of information can be selected when selecting information in the knowledge DB 4 as described above, the fact that a plurality of pieces of information exist at the time of presenting the information is displayed so that the user can freely select the information. , The probability that the user can refer to the information desired is increased.

[0047]

As described above, according to the present invention, it is possible to reuse a process required for solving a problem in the past by another user, and to remove only a necessary portion from other users' trial and error. Referencing has the advantage that efficient access can be achieved for the user's own search purpose. In particular, of the history information to be referred to, by extracting only a portion that is relatively highly likely to be truly effective and sequentially presenting it to the user, there is a high possibility that the information desired by the user can be accessed. There are advantages. When appropriate information cannot be accessed due to the division of information, not only the partial reference information but also the history information can be used to sequentially show the entire information, providing the user with the necessary flexibility. Can respond.

[Brief description of the drawings]

FIG. 1 is a system configuration diagram according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a main part of the embodiment of the present invention.

FIG. 3 is a flow chart according to the embodiment of the present invention.

FIG. 4 is a diagram showing a state of information search.

FIG. 5 is a diagram showing an example of an HTML source file.

FIG. 6 is a diagram illustrating an example of information division.

FIG. 7 is a diagram illustrating an example of a knowledge DB.

FIG. 8 is a diagram showing a display example of a result.

FIG. 9 is a diagram illustrating an example of accumulation in a knowledge DB.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 user 2 reference history storage unit 3 relay server 4 knowledge database (DB) 5 information 6 WWW (World Wide Web) 7 WWW browser 30 search purpose extraction unit 31 reference history acquisition unit 32 information acquisition unit of answer candidate 33 information division unit 331 Unnecessary information exclusion means 34 Knowledge accumulation means 341 Appearance frequency recording means 35 Search purpose search means 36 Information extraction means 37 Information display means

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G06F 15/419 320

Claims (9)

[Claims] An information reference method for efficiently performing information reference using information possessed by another user when referring to a user's purpose from a vast amount of information. A step of finding a search purpose for the information to be referred to; a step of obtaining a reference history for information that can be a candidate for the answer from a huge amount of information sources for the search purpose for the information; Acquiring information that can be an answer to the search purpose from the information, dividing the information that can be the answer, storing the divided information in the knowledge database in the order of appearance together with the search purpose, Searching for the closest search objective from among the search objectives stored in the knowledge database when the objective to be referred to exists in the knowledge database; An information reference method, comprising: a step of extracting information with an order that matches a searched purpose; and a step of displaying the information with the extracted order in the order. 2. From information divided according to the subject,
2. The information reference method according to claim 1, further comprising a step of excluding unnecessary information, and a step of storing only information left by said excluding step in said knowledge database. 3. A step of recording the appearance frequency of the information stored in the knowledge database, and a step of recording a new purpose to be referred to, the information being stored in the knowledge database. A step of performing a search by assigning a weight to the object having a high appearance frequency when searching for the closest search object from the search objects.
Information reference method described. 4. A step of, when extracting information with an order that matches a search purpose, confirming whether a part of the information corresponds to a part of information with another order, 2. The information referencing method according to claim 1, further comprising the step of, if a result of the matching is found, displaying the result simultaneously or sequentially as one of the options when displaying the result. 5. An information referencing device for efficiently referencing information intended by a user from among a vast amount of information using information possessed by another user. Means for finding a search purpose for the information to be referred to, means for acquiring a reference history for information that can be a candidate for the answer from a vast information source for the search purpose for the information, Means for acquiring information that can be an answer to the search purpose from the search means, means for dividing the information that can be the answer, means for storing the divided information in the knowledge database in the order of appearance together with the search purpose, When there is an object to be referred to, a means for searching for the nearest search object from among the search objects stored in the knowledge database, and an order matching the searched object, An information referencing device comprising: means for extracting accompanying information; and means for displaying information with the extracted order in the order. 6. From information divided according to the subject,
6. The information reference apparatus according to claim 5, further comprising: means for excluding unnecessary information; and means for accumulating only information left in the exclusion step in the knowledge database. 7. A means for recording the frequency of appearance of the information stored in the knowledge database and a purpose to newly refer to the information, the information being stored in the knowledge database. 6. The information reference apparatus according to claim 5, further comprising: means for performing a search by weighting the one having a high appearance frequency when searching for the closest search objective from the search objectives. 8. The means for extracting information with an order suitable for the searched purpose includes means for confirming whether a part of the information matches a part of the information with another order, 6. The method according to claim 5, wherein the means for displaying the information includes means for simultaneously or sequentially displaying one of the options at the time of displaying the result if the matching information is found as a result of the confirmation. Information reference method described. 9. A method for realizing an information referencing device for efficiently performing information referencing using information possessed by another user when referring to an object intended by a user from a vast amount of information. A process for finding a search purpose for information that a user wants to refer to, and a process for finding information that can be a candidate for an answer from a vast amount of information sources for the search purpose for the information. A process of acquiring a reference history, a process of acquiring information that can be an answer to the search purpose from the acquired reference history, a process of dividing the information that can be the answer, and the appearance of the divided information together with the search purpose. The process of storing in the knowledge database in order, and the search purpose stored in the above knowledge database when there is a new purpose to be referenced The process of searching for the closest search purpose, the process of extracting information with an order that matches the searched purpose, and the process of displaying the information with the extracted order in that order are performed by the computer. A storage medium storing an information reference program storing a program to be executed.