JP2000112968A - Information retrieval supporting device and information retrieval supporting program storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To supply a keyword, which can be used for retrieval and on which the retrieval intention of a user is reflected, to the user to an information retrieval indication device indicating the keyword considered to be suitable for permitting the user to use for the retrieval of a desired document to the user, and also on an information retrieval program storage medium where an information retrieval program for operating a computer as the information retrieval indication device is stored. SOLUTION: The device is provided with an association table storage part 2 where an association table filled with information on the presence or absence of association property between respective keys constituting a key group and respective pieces of association information constituting an association information group is stored, a key input part 1 inputting the keys, an associated key extraction part 3 receiving the input of the key from the key input part, referring to the association table stored in the association table storage part and extracting association information associated with the input key being the inputted key and an associated key being the key different from the associated input key, and an associated key output part 4 outputting the associated key extracted by the associated key extraction part 3.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information retrieval / presentation apparatus for presenting a user with a key such as a keyword which is considered suitable for use in searching for a document or the like desired by the user.
And an information search program storage medium storing an information search program for causing a computer to operate as the information search and presentation device.

[0002]

2. Description of the Related Art In recent years, the development of computer technology such as database technology and network technology found on the Internet has led to a wide variety of document information (hereinafter simply referred to as "documents") stored in databases distributed on a network. ) Can be used without the user being aware of the location of the database. However, it is unlikely that the user will need all of a wide variety of documents, and the user needs to perform an operation of selecting a document having useful information from the large number of documents.

At present, a large number of search engines having a keyword search function are provided as a system for supporting a task of selecting useful documents from a large number of documents distributed on the Internet. In the keyword search, a document in which a keyword given by the user appears from a group of available documents is extracted and presented. Therefore, a user needs to input a keyword suitable for performing a keyword search for a document including desired information into the system. However, it is burdensome for the user to come up with an appropriate keyword necessary for obtaining the target document. Therefore, a system that recommends a keyword to be input to a user is useful.

As a conventional technique for supporting a user's keyword search by recommending a keyword, for example, when a certain document or document group is found as a result of performing a search using a certain keyword in the past, the document or the document group is searched. Based on the assumption that the words appearing in the document are words related to the keyword, and based on the evaluation method using the frequency of appearance of many words appearing in the document or in the document group, as a keyword related to the keyword There is known a technique of extracting usable related keywords, and recommending related keywords extracted when a user inquires about keywords related to the keywords. As another example, a thesaurus in which synonyms and similar words are associated with each other is prepared, and when a user enters a certain keyword and inquires about a keyword related to the keyword, the user refers to the thesaurus, There is also known a technique of recommending a keyword composed of a synonym or a similar word of the input keyword.

[0005]

However, among the above-mentioned prior arts, in the case of the technique of recommending a keyword obtained based on the frequency of appearance of a word appearing in a document retrieved in the past, the intention of the document provider is In other words, words used frequently by the writer who created the document are extracted as recommended keywords, and the search intention of the user is not reflected therein.

[0006] Among the above-mentioned conventional techniques, a technique of recommending a keyword composed of synonyms and similar words with reference to a thesaurus is an evaluation of words based on the relationship between words in linguistics. Search intent is not reflected there.

[0007] On the other hand, the search intentions of the user are various, and the evaluation of the word as a candidate for the recommended keyword may differ for each search of the user. However, in the above-described conventional technology, there is a problem that a keyword cannot be recommended in accordance with an evaluation reflecting various search intentions of a user because a difference in search intentions of the user is not reflected.

In view of the problem that a user cannot always imagine a key such as a keyword suitable for searching for desired information, the present invention extracts a key for information search by a new method. To provide an information search support device to be recommended to a user, and an information search support program storage medium storing an information search support program for causing a computer to operate as such an information search support device. .

[0009]

A first information retrieval support apparatus of the information retrieval support apparatus of the present invention for achieving the above object comprises a key forming a predetermined key group and a predetermined related information group. A related table storage section in which a related table containing information on the presence or absence of relevance between each related information is stored, a key input section for inputting a key, receiving a key input from the key input section, A related key extracting unit that refers to a related table stored in a related table storage unit and extracts a related key that is a key different from the input key that is further related to related information related to an input key that is an input key. , And a related key output unit that outputs the related key extracted by the related key extraction unit.

Here, the key is information that plays a role as a trigger for information retrieval. A word used as a keyword in a keyword search is a key. In the present invention, if the information serves as a trigger,
Not only words but also various forms of information such as documents, users, images, sounds, and the like, and sets thereof can be treated as keys, and the key in the present invention is a general term for them.

The related information is information related to key information. For example, a document in which a key appears is related information related to the key based on the appearance relationship, and a user using the key is related information related to the key based on the usage relationship. Thus, in the present invention, various information can be related information for the key.

An association table is a table showing the correspondence between keys and related information for the keys. The association table contains information indicating the presence or absence of association between each key and each piece of related information. Is included in the association table in the embodiment described below. In addition to the presence / absence of association, information (e.g., here, an evaluation value indicating the degree of association between each key and each piece of related information) , Referred to as “key information relevance”)
May be included. In the following, when it is necessary to distinguish between a related table that includes information indicating the presence / absence of a relationship but does not include the relevance between key information and a related table that includes the relevance between key information, each of them is referred to as an “evaluation value”. Tables with no evaluation value "and" association tables with evaluation values ".

[0013] The degree of association between key information is a concept of an evaluation value obtained by evaluating the degree of association between a key and related information as described above. The evaluation value obtained by evaluating the degree of association is referred to herein as “key association degree”.

In the first information retrieval support apparatus of the present invention, a relation table created in advance is used, and a user can receive a recommendation of a relation key for an input key.

[0015] The device form of the first information search support device of the present invention may be a stand-alone type device which is isolated from other devices, or transmits and receives information via a communication path. It may be a type of device.

In the case of the former stand-alone type device, for example, a keyboard and a mouse correspond to the key input unit according to the present invention, and the image display device corresponds to the related key output unit according to the present invention.

In the latter type, that is, when the information retrieval support device is connected to a communication path, a receiving section for receiving a key transmitted via the communication path corresponds to a key input section. The transmitting unit that transmits the related key extracted by the related key extracting unit to the communication source that transmitted the input key received by the key input unit corresponds to a related key output unit.

Here, in the first information retrieval support apparatus of the present invention, the association table storage unit indicates not only the presence or absence of association but also the degree of association between each key and each item of associated information. It stores an association table in which the degree of association between key information is described.
With reference to the relation table stored in the relation table storage unit, the inter-key relation degree indicating the degree of relation between the input key and the relation key extracted by the relation key extraction unit in relation to the input key is calculated. When a plurality of related keys are extracted in relation to the input key, a relationship between the plurality of related keys is determined according to the inter-key relevance determined by the related degree calculation unit. It is preferable to include a related key order determining unit that determines the order of relevance, and a related key order output unit that outputs information indicating the order of the related keys determined by the related key order determining unit.

With such a configuration, the user can not only receive recommendation of a related key with respect to an input key designated by the user, but also receive order information indicating the superiority or inferiority of the recommended related key. This makes it easy to determine which key to use as a search key.

Further, in the first information retrieval support device, the input key and the already extracted input key which are further related to the related information further related to the related key extracted by the related key extracting unit are stored in the related key extracting unit. It is also a preferable embodiment to provide a control unit for extracting related keys other than the related keys.

In this case, a related key that is further related to the once extracted related key is transitively obtained, so that the user can be recommended more related keys.

Further, in the first information search support device, the key input unit inputs a plurality of keys, and the information search support device further includes a plurality of input keys input from the key input unit. Control for causing a combination condition input unit for inputting a combination condition for a combination key and a related key extraction unit to extract a plurality of input keys input from the key input unit and a relevant key matching the combination condition input from the combination condition input unit It is preferable to include a part.

In the case of this configuration, as a result of the provision of the above-mentioned combination condition input section and the above-mentioned control section, it is possible to extract a related key from a plurality of input keys and a combination condition obtained by combining them. It becomes possible to give various related key recommendation instructions.

Further, in the first information retrieval support device, the association table storage section stores a plurality of types of association tables, and the information retrieval support device further comprises:
A related table designating unit that designates one or more related tables to be searched, and a related key extracting unit that is designated by the related table designating unit of a plurality of types of related tables stored in the related table storage unit. 1
It is preferable to provide a control unit for extracting a related key with reference to one or more related tables.

In the case of this configuration, the user can specify a desired relation table among a plurality of types of relation tables,
It becomes easier to collect related keys that are useful for the user. In selecting a key to be used for information retrieval from the specified related keys, the selection is performed in consideration of what kind of information the specified related key is obtained through. Can be performed, and a key effective for searching for desired information can be selected.

Further, in the first information retrieval support apparatus, an output related key specifying unit for specifying a related key restriction condition for restricting the related key extracted by the related key extracting unit; Preferably, a control unit is provided for extracting a related key in a range satisfying the related key restriction condition specified by the output related key specifying unit.

Here, the related key restriction condition is not limited to a specific type of condition. For example, the related key restriction condition further includes the range of the degree of relevance when the related table with the evaluation value is adopted, and the extracted related key. The number of repetitions in a mode of repeating a related key extraction cycle of extracting a related key may be used as a related key restriction condition. Furthermore, a specific key (unrelated key) is designated, and when the unrelated key is extracted, a request for stopping the extraction of the related key is required, or the unrelated key is designated and the unrelated key is extracted. Even if it is performed, a request to remove the key from the related key may be set as the related key restriction condition.

In the first information retrieval support apparatus according to the present invention, when the output related key designating section for designating the related key limiting condition is provided, the likelihood that a related key suitable for the purpose of the user is recommended increases. This makes the device more valuable.

Further, in the first information retrieval support apparatus of the present invention, an update information input section for inputting update information for updating a relation table stored in the relation table storage section; It is preferable to include a related table updating unit that updates the related table stored in the related table storage unit based on the input update information.

In this case, it is preferable that the update information input section includes means for generating update information based on a history of past information collection.

[0031] A flexible information retrieval support system is constructed by allowing the related table to be updated afterwards, instead of being fixed once created. In particular, if a means for generating update information based on the history of past information collection is provided, an association table reflecting the user's search tendency is automatically created, and this information search support device is more useful. It will be.

Here, in the first information search support device, the related key extracting unit extracts related information related to the input key with reference to a predetermined related table, and further relates to the extracted related information. The related key may be extracted by referring to the same related table as the predetermined related table. Alternatively, the related table storage unit stores a predetermined first related table, The second related table having a key group having a different genre from the key group of the first related table and a related information group having the same genre as the related information group of the first related table. , The related key extracting unit extracts related information related to the input key by referring to the first related table, and further extracts a related key related to the extracted related information to the second related table. May be extracted with reference to There may be a combination thereof both.

Further, the information search support device of the present invention is
A device which is connected via a communication path to an apparatus having an association table storing section in which the above-mentioned association table is stored, and performs information retrieval support by referring to the association table stored in the association table storage section of the apparatus. It may be. Here, among the information search support devices of the present invention, the information search support device thus configured is referred to as a second information search support device of the present invention.

That is, the second information retrieval support apparatus according to the present invention describes the information on the presence or absence of relevancy between each key constituting a predetermined key group and each related information constituting a predetermined related information group. A key input unit for inputting a key, which is connected via a communication path to a device provided with an association table storage unit storing the associated association table, receiving the key input from the key input unit, and storing the association table With reference to the related table stored in the section, further related to the related information related to the input key that is the input key,
An information retrieval support device comprising: a related key extracting unit that extracts a related key that is a key different from an input key; and a related key output unit that outputs a related key extracted by the related key extracting unit.

According to the second information search support device of the present invention, a plurality of information search support devices are connected to the device provided with the association table storage unit, and the plurality of information search support devices store the related table storage unit. Can be shared.

Here, only the basic form of the second information retrieval support apparatus of the present invention has been described here, but this is for the purpose of avoiding repeated explanation, and the related table storage unit is configured as another apparatus. Various forms of the first information search support device of the present invention are directly applicable to various forms of the second information search support device of the present invention except that the device is connected to the device by a communication path. I do.

Further, the information search support program stored in the information search support program storage medium of the present invention, which achieves the above object, receives input keys serving as a search trigger and forms a predetermined key group. A related information extracting means for extracting a related information related to an input key by referring to a related table in which information of presence or absence of a relationship between each key and each related information configuring a predetermined related information group is provided; A related key extracting unit that extracts a related key that is different from the input key and that is further related to the related information extracted by the related information extracting unit.

When the information retrieval support program operates in a computer, the related information extracting means and the related key extracting means are integrated with each other.
It will function as a related key extraction unit in the information search support device of the present invention.

The information search support program according to the present invention will be described only in its basic form, but this is merely for the purpose of avoiding repeated description. In addition to the information search support program of the basic form described above, various types of information search support programs for realizing various forms of the first information search support apparatus or the second information search support apparatus described above are included. In particular, the information search support program according to the present invention may include an association table creating means for creating an association table.

[0040]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the basic concept of the present invention will be described first, and then the embodiments of the present invention will be described.

Table 1 conceptually shows a relation table without an evaluation value.

[0042]

[Table 1]

The relation table shown in Table 1 indicates that there is a relation between the key adjacent to the left and right and the related information, that is, the i-th key i and the related information i.

Table 2 conceptually shows an association table with evaluation values.

[0045]

[Table 2]

In the association table shown in Table 2, n keys are listed in the row direction and m pieces of related information are listed in the column direction.
The degree of relevance between key information, which indicates the degree of relevance between the key i and the related information k, is represented by an evaluation value ik.

Here, a related table without the evaluation values shown in Table 1 can be shown in Table 2. In this case, the evaluation value is represented by two values, for example, “1” when there is relevance and “0” when there is no relevance.

Table 3 is a specific example of an association table without an evaluation value.

[0049]

[Table 3]

Table 3 is an association table in which a keyword, that is, a word is used as a key and a document different from the key is used as related information. In Table 3, the keyword "Apple" has a relevance with the information "Document 1" based on, for example, a relationship "Keyword and a document searched using the keyword".

Table 4, Table 5, Table 6, Table 7, and Table 8 are specific examples of evaluation-related tables.

[0052]

[Table 4]

[0053]

[Table 5]

[0054]

[Table 6]

[0055]

[Table 7]

[0056]

[Table 8]

Tables 4, 5 and 6 use a keyword, that is, a word as a key, and are different from the key.
It is an association table in which a user and a search formula are associated information. On the other hand, Table 7 is a diagram of a relation table showing a relation in which the relation between the keyword as the key of the relation table shown in Table 4 and the document as the related information is reversed. The keys and related information are role names, and as shown in Table 7, there are cases where roles can be exchanged. Table 8 shows an association table in which a keyword does not appear as information and a user is a key and a document is information related to the key.

Tables 4, 5, 6, and 8 will be described in more detail as follows.

As in Table 3, Table 4 has a relationship of a keyword as a key, a document as related information for the key, and a relationship between the key and related information as “a keyword and a document searched using the keyword”. And an evaluation value-equipped association table having an evaluation value indicating a degree of relevance indicating “usefulness of information held by a document with respect to a keyword”. For example, in Table 4, when the degree of association between the key “Apple” and the related information “Document 1” is “10”, the usefulness of the information of “Document 1” expressed by the keyword “Apple” is determined. This indicates that the evaluation value to be represented is “10”.

Similarly, Table 5 shows that a keyword is a key, a search formula is related information for the key, a relationship between the key and the related information is a relationship of “keyword and a search formula searched using the keyword”, and It is an association table with an evaluation value, where the degree of association is an evaluation value indicating “frequency of use of search formula”. Table 6 shows that a keyword is a key, a user is related information for a key, a relationship between a key and related information is a relationship of “keyword and a user using the keyword”, and a degree of relevance is “keyword by user”. It is an association table with an evaluation value, which is an evaluation value representing “frequency of use”. Table 8 shows that
The key is the user, the document is the related information for the key, the relationship between the key and the related information is "user and the document used by the user", and the degree of relevance is "document frequency".
Is an association table with an evaluation value, which represents an evaluation value.

FIG. 1 is an explanatory diagram of a related key.

A related key for a key is a key having related information common to the key. For example, in FIG. 1, when the keys k1 and k2 are both keys having the document d as the related information, the keys k1 and k2 are mutually related keys via the document d which is the common related information (hereinafter, “related Key relationship). A key having a transitive relationship obtained by combining two or more different related key relationships is called an indirect related key. When a related key and an indirect related key are not distinguished, they are simply referred to as a related key.

FIG. 2 is an explanatory diagram of a basic principle for extracting a related key according to the present invention.

The key information relevance appearing in the relevance table in FIG. 2 represents the degree of relevance by a non-negative value.
The presence or absence of the relevance is determined based on whether or not the threshold exceeds a preset threshold (set to 0 in the following description). In the present invention, the basic process of finding a related key for a key proceeds by the following steps or by repeating the entire steps.

Step 1: Key 1 is designated as a key for which a related key is recommended.

Step 2: Information 1 and 2 are selected as information related to key 1 in which the degree of association between key information, which is an evaluation value for key 1, exceeds a threshold.

Step 3: The related information 1 and 2 are related.
That is, the key 1 previously designated as the key whose degree of association between key information, which is the evaluation value of the related information 1 and 2, exceeds the threshold value.
Are selected.

Step 4: Keys 2 and 3 are designated as related keys for key 1.

The related keys 2 and 3 corresponding to the key obtained as described above are keys to which the same related information as the related information to which the key 1 is related is related. The related key obtained in this way, and the related key including the indirect related key obtained when the above processing is repeatedly applied using the related key, are the recommended keys to the user with respect to the original input key. Becomes

FIG. 3 is a diagram showing an embodiment of the device configuration of the information search support device of the present invention.

The information retrieval support device 100A shown in FIG.
Is an information search support device constituted by a stand-alone type computer system, which includes a CPU 101 for executing various programs, a memory 102 storing various programs, a keyboard and a mouse, etc. The apparatus includes an input device 103 for giving various instructions to the device 100A, an output device 104 for displaying various information and providing information to a user, and a database 105 storing a relation table according to the present invention.

In this embodiment, an information search support program for operating this computer system as an information search support device is stored in a portable storage medium 200 such as an optical disk or a magneto-optical disk and supplied. The data is loaded from the memory 200 to the memory 102 in the computer system, whereby the computer system operates as an information search support device.

The information retrieval support device 100A shown in FIG.
When compared with the basic mode of the present invention, the input device 103 corresponds to the key input unit of the present invention, the output device 104 corresponds to the related key output unit of the present invention, and the database 105 corresponds to the related table storage unit of the present invention. Correspondingly, a program stored in the memory 102 and the CPU 10 for operating the program
The related key extraction unit according to the present invention is configured by a combination with the key combination (1).

FIG. 4 is a diagram showing another device configuration of the information search support device of the present invention. The difference from the information search support device 100A shown in FIG. 3 will be described.

The information retrieval support device 100B shown in FIG. 4 includes a reception device 106 and a transmission device 107 connected to the communication path 300.
OB is connected to a plurality of search devices 400_1,..., 400_n via the communication path 300.

Here, the receiving apparatus 106 receives an input key from one of the plurality of search apparatuses 400_1,..., 400_n (here, the search apparatus 400_1) via the communication path 300. When the transmission device 107 receives the input key,
Plays a role of sending out a related key corresponding to the input key to the communication path 300 toward the search device 400_1 that transmitted the input key. That is, in the case of the information search support device 100B shown in FIG. 4, the receiving device 106 corresponds to the key input unit according to the present invention, and the transmitting device 107 corresponds to the related key output unit according to the present invention.

FIG. 5 is a diagram showing another device configuration of the information search support device of the present invention.

The information search support device 100C shown in FIG.
And the database 500 are connected by a communication path 300.

The information search support device 100
A plurality of information search support devices 100 having the same configuration as C
, 100N are shown, and the information search support device 1
A plurality of information search support devices 100C and 10 including 00C
, 100N share the database 500.

The information search support device of the present invention is shown in FIGS.
The information search support device of any one of the above-described embodiments is applicable. In each of the examples shown below, it is necessary to distinguish which of the device forms of FIGS. I will explain without doing it.

FIG. 6 shows a first example of the information search support apparatus of the present invention.
It is a functional block diagram of an embodiment.

In the figure, a key input unit 1 is a functional unit having key input means for inputting a key to present a related key used by a user for a search.

The association table storage unit 2 is a functional unit having an association table storage unit for storing an association table.

The related key extracting unit 3 receives an input key input by the user from the key input unit 1 and extracts a related key corresponding to the received input key from the related table stored in the related table storage unit 3. This is a functional unit having an extracting unit.

The related key output unit 14 is connected to the related key extracting unit 3
Is a functional unit having related key output means for receiving the related key obtained in (1) and presenting the received related key as a related key to the input key.

FIG. 7 is a flowchart showing a procedure of processing performed by the information search support device shown in FIG.

In the figure, a process s1 is a process performed by the key input unit 1 in FIG. 6, processes s2 and s3 are processes performed by the related key extraction unit 2 in FIG. 6, and a process s4 is a related key output unit 4 in FIG. Represents the processing to be performed.

Process s1: A key to which a related key is presented is input.

Process s2: Information related to the input key obtained in process s1 is obtained using the relation table stored in the relation table storage unit 2 in FIG.

Process s3: A key group further related to the information related to the input key obtained by process s2 is shown in FIG.
Is obtained by using the related table stored in the related table storage unit 2 in.

Process s4: The key group obtained in process s3 is presented as a related key to the key input by the user.

As a result of the above processing, the user can receive a recommendation of a related key for the key designated by the user.

The first embodiment shown in FIGS. 6 and 7 will be described more specifically.

FIG. 8 is an explanatory diagram showing a specific example of the first embodiment.

Now, the related table shown in Table 3 without the evaluation value is stored in the related table storage unit 2 in FIG. 6, and the user inputs the key (keyword) “calculator” in FIG. It is assumed that the input is made to the unit 1. Here, as described above, when the input source is a human (user), the input to the key input unit 1 is input via an input device such as a keyboard or a mouse, or may be performed by another computer system. For example, a method of receiving as an input signal through a line is adopted.

At this time, the processing shown in FIG. 8 is performed in the related key extracting unit 3 in FIG. 6, which is a functional unit having related key extracting means. Hereinafter, steps of the processing illustrated in FIG. 8 will be described.

Step 1: "Calculator" is designated as an input key.

Step 2: “Document 1” and “Document 2” are selected as related information related to the key “computer”.

Step 3: related information related to the key "computer" As keys related to the "document 1", "Apple" and "database" different from the key "computer" are used, and related information related to the key "computer". Similarly, “Network”, which is different from the key “Computer”, is selected as the key related to “Document 2”.

Step 4: "Apple", "database", and "network" are designated as related keys for the key "computer".

The related keys “Apple”, “database”, and “network” for the key “computer” obtained from the above are output from the related key output unit 4 in FIG. 6 as the recommended keys for the key “computer”. Is output.

The related keys “Apple”, “database”, and “network” for the key “computer” are related keys extracted using the related table shown in Table 3, which reflects the evaluation of the user.

The output from the related key output unit is output via an image output device such as a CRT when the output destination is a human (user), and is output through a line in another computer system. For example, a method of transmitting the information is used.

In the conventional method, a user is limited to a method of expressing an intention to be input when receiving a recommendation, and is limited to a method of expressing a keyword, that is, a character.
Intentional expression cannot be performed in various expression forms not limited to character strings. On the other hand, the present invention not only recommends the related keyword for the keyword using the association table shown in Table 3 as described with reference to FIG. 8, but also the association table shown in Table 7 or Table 8. By using a document or user that is not a keyword as a key, the same applies to recommending a key, document, or user that is related to the input key, that is, the input document or user. It is possible in the framework. Furthermore, the present invention is characterized in that by preparing various association tables, it is possible to recommend using keys expressed in various expression forms not limited to character strings such as voices and images.

In the following, using the relation table shown in Table 8 in which the key is the document user, the related information is the referenced document, and the degree of association between key information is the number of times the document user has referred to each document, the designated key is used. The following describes the procedure for recommending the document user serving as the related key to the document user “Sato”.

FIG. 9 is a diagram schematically illustrating the procedure.

However, each table in FIG. 9 is based on the relation table with evaluation values shown in Table 8 when the degree of association between key information (ie, the number of times a document user refers to each document) is greater than 0.
This is an association table without an evaluation value, which is determined to have a corresponding relationship (that is, association) between the key (document user) and the information (document). For example, in the association table shown in Table 8,
As for the key “Sato”, only the information “Document 4” has a value “4” larger than 0 as the degree of relevance between key information (ie, the number of times a document user refers to each document). In the related table without the evaluation value, only the information “Document 4” appears as a pair with respect to the key “Sato”.

Step 1: The document user "Sato" is designated as an input key.

Step 2: “Document 4” is selected as related information related to the key “Sato”.

Step 3: "Suzuki" and "Tanaka" different from the key "Sato" are selected as keys related to the related information "Document 4" for the key "Sato".

Step 4: "Suzuki" and "Tanaka" are designated as related keys to the key "Sato".

The related keys "Suzuki" and "Tanaka" for the key "Sato" obtained as described above are the keys recommended for the key "Sato", as in the previous example, the related key output unit 4 in FIG. Output. The related keys “Suzuki” and “Tanaka” corresponding to the key “Sato” have a relationship of a related key extracted in the relationship that the document user “Sato” refers to the same document as the document referred to.

In the above description with reference to FIG. 9, step
The association table for obtaining information related to the input key in 1 and 2 matches the association table for obtaining a key from information in steps 3 and 4. However, their association tables need not match. For example, in the example shown in FIG.
By using the relation table shown in Table 8 as the relation table used in 3 and 4, the relation key for the input key "calculator" can be obtained not from the keyword but from the user.

FIG. 10 is a functional block diagram of a second embodiment of the information search support device of the present invention.

The difference between the functional block and its configuration shown in FIG. 10 and the functional block and its configuration of the first embodiment shown in FIG. 6 described above is that the following functions and functional blocks are added. .

1) The relation table stored in the relation table storage unit 2 shown in FIG. 10 includes, in addition to the information of the relation table in FIG. 6, information on the degree of relevance between keys and information, which is the degree of relevance between keys and information. Having.

2) The relevance calculating section 5 includes the relevant key extracting section 3
Is a functional unit having relevance calculating means for receiving an input key and a relevant key from the relevant table and using a relation table with an evaluation value stored in the relevant table storage unit 2 to calculate the relevance of each relevant key to the input key.

3) The related key order determining unit 6 receives the related key group with the related degree from the related degree calculating unit 5, and determines the order of the related key according to the relative degree of the related degree of each related key. This is a functional unit having a rank determining unit.

4) The related key rank output section 7 is a functional section having related key rank output means for presenting the rank determined by the related key rank determination section 6. FIG. 11 is a flowchart illustrating a procedure of a process performed by the information search support device illustrated in FIG.

Processing s1, s2, s3, s in FIG.
8 are common to the processes s1, s2, s3, and s4 in the first embodiment described with reference to FIG. The processes s4, s5, and s6 are the processes performed by the relevance calculator 5 in FIG. 10, the process s7 is the process performed by the related key order determination unit 6 in FIG. 10, and the process s9 is the related key order output unit in FIG. 7 represents the processing performed.

Process s1: A key to which a related key is presented is input.

Process s2: Information related to the input key obtained in process s1 is obtained using the relation table stored in the relation table storage unit 2 in FIG.

Process s3: A key group to which information related to the input key obtained by process s2 is related is obtained by using the relation table stored in the relation table storage unit 2 in FIG.

Process s4: The degree of relevance between the input key obtained in process s1 and the information related to the input key obtained in process s2 is determined using the relation table stored in the relation table storage unit 2 in FIG. Ask.

Process s5: The degree of association between the information related to the input key obtained in process s2 and each key included in the key group related to the information related to the input key obtained in process s3 is It is determined using the association table stored in the association table storage unit 2 in FIG.

Process s6: From the two degrees of association obtained in processes s4 and s5, the input key obtained in process s1 and each key included in the key group related to the information related to the input key obtained in process s3 And the degree of relevance between them.

Process s7: The order of each related key with respect to the input key obtained in process s1 is determined according to the degree of relevance between keys obtained in process s6.

Process s8: The key group obtained in process s3 is presented as a related key to the key input by the user.

Process s9: The order of each related key obtained in process s7 is presented.

As a result of the above-mentioned processing, the user can receive recommendation of related keys with respect to the keys set by the user and rank information indicating the superiority or inferiority of the related keys.

Here, how to determine the degree of association in the present embodiment will be described.

Assume that the relation table shown in Table 4 is stored in the relation table storage unit 2 in FIG. 10, and that a key (keyword) “calculator” is input from the user to the key input unit. In the present embodiment, the following expression (1) is used as a relevance calculating means for calculating the inter-key relevance in the relevance calculator 5 shown in FIG. The expression (1) means that the degree of association between the key k1 appearing in the related table R of interest and the related key k2 is the evaluation value of each key for each piece of related information appearing in the related table R of interest. However, when another related key is used, the value is set higher.

[0133]

(Equation 1)

Here, R: a relation table to be selected, a ′: a key included in a set A of keys appearing in the relation table R, a ″: a relation included in a set A of keys in the relation table R for the key a ′ Key, b: related information included in set B of related information appearing in related table R, v _R (a, b): key a and related information b in related table R
W _R (a, b): evaluation table for evaluation value v _R (a, b) for evaluation value v _R (a, b)
Is the weight of the related information b for the key a in

[0135]

(Equation 2) And the value given by

At this time, in the information retrieval support apparatus shown in FIG. 10, in the same manner as in the first embodiment shown in FIG. 6, the related key corresponding to the key "computer" is stored in the related key extracting section 3 having related key extracting means. Ask in. The related keys for the key “computer” obtained as a result are “Apple”, “database”, and “network” as in the example shown in FIG. When the inter-key association degree for the key “calculator” and each of these related keys is obtained according to the equation (1),
They are 0.4, 0.4, and 0.2, respectively.

That is, for the key “calculator”,
Since the evaluation values of “document 1”, “document 2”, and “document 3” are 10, 15, and 0, respectively, for the document, using equation (2), w _R (computer, document 1) =） 10 / (10 + 15 +
0)｝ = 0.4 With respect to the key “database”, the evaluation values of “document 1”, “document 2”, and “document 3” are 20, 0, 0, respectively. Using equation (2), w _R (database, document 1) = ｛20 / (20 + 0 +
0)｝ = 1.0 Therefore, according to equation (1), the degree of inter-key association between the key “computer” and the key “database” is r _R (computer, database) = 0.4 × 1.0 = 0 .
4 The same applies to other inter-key association degrees.

[0138] In the expression (1), "Σ" representing addition is shown. For example, in the example of Table 4, the route from the key "computer" to the key "database" intervenes "document 1". In the case where there is only a route that has been made to exist, but there is another route that is not shown in Table 4, for example, a route that has made “document 4” besides the route that has made the “document 1” (that is, the key “computer”) Is related to Document 4,
(If "Document 4" is also related to the keyword)
, The relevance of the two routes is added.

FIG. 12 is a functional block diagram of a third embodiment of the information search support device of the present invention.

The difference between the functional block and its configuration shown in FIG. 12 and the functional block and its configuration of the first embodiment shown in FIG. 6 is that the following functional blocks are added.

The control unit 8 shown in FIG. 12 has related key extraction control means for extracting the related key corresponding to the related key by inputting the related key extracted by the related key extracting unit 3 to the related key extracting unit 3 again. It is a functional unit.

FIG. 13 is a flowchart showing a procedure of processing performed by the information search support device shown in FIG.

The flow of the process performed by the information search support device shown in FIG. 12 is basically the same as the process in the first embodiment shown in FIG. 6, and processes s1, s4, and s4 in FIG.
Steps s5 and s7 are the same as steps s1, s2, s3 and s4 shown in FIG. 7 in the first embodiment shown in FIG. 6, respectively. The difference is that the control unit 8 includes processes s2, s3, and s6 performed by the control unit 8 in FIG.

Hereinafter, processes s2 and s unique to this embodiment will be described.
3, s6, and other processes are abbreviated. However,
In FIG. 13, the “key of interest” has a meaning of a key for which a related key is to be obtained. Hereinafter, the term “key of interest” is used in the meaning here.

Process s1: Same as process s1 in FIG.

Process s2: The input key is set as a key of interest on which a related key is obtained.

Process s3: If the key of interest can be set, the process moves to process s4. If not, the process moves to step s7.

Processes s4 and s5: Processes s2 and s2 in FIG.
Same as s3.

Process s6: One of the related keys extracted in process 5 is selected, and is set as a target key from which the related key is obtained.

Process s7: Same as process s4 in FIG.

Processing s Performed by Control Unit 8 in FIG.
As a result of the existence of 2, s3, and s6, in the information search support device of the present embodiment, a process of obtaining a related key for a related key is added to the process performed in the basic embodiment shown in FIG. 6, and the process is repeated. As a result, the effect that more related keys are required to be transitively obtained is obtained. As a result, the user can receive a recommendation of a related key for a key set by the user and a related key for a related key that is transitively required.

FIG. 14 is a schematic diagram showing the latter half of a specific example of the processing flow in the third embodiment described with reference to FIGS.

Assuming that a situation similar to that described with reference to FIG. 8 has been set, the related key extracting unit 3 in the third embodiment shown in FIG. 12 applies the related key extracting operation once. It is assumed that the same result as in FIG. 8 is obtained. At this time, by operating the related key extracting unit 3 again,
The process of obtaining an indirectly relevant key that is a related key that is a related key to a related key will be specifically described with reference to FIG.

FIG. 14 shows that “Apple”, “database”, and “network” are obtained as the related keys to the key “computer” (via steps 1 to 4 shown in FIG. 8), and these are set as the attention keys again. From the time point (step 5), through a process similar to the process described with reference to FIG. 8, via the related information “document 3” for the related key “network” (step 6), the related key “ A "banana" is determined (step 7), and the key "banana" is selected as an indirectly related key to the input key "computer" (st7).
ep8) represents the process.

FIG. 15 is a schematic diagram showing the latter half of another specific example of the processing flow in the third embodiment described with reference to FIGS.

Similarly, assuming that a situation similar to the situation described with reference to FIG. 9 is set, and the related key extracting operation is applied once by the related key extracting unit 3, a result similar to that of FIG. 9 is obtained. Is obtained. At this time, the process of obtaining the indirect related key which is the related key for the related key by operating the related key extracting unit 3 again will be specifically described with reference to FIG.

FIG. 15 shows, as a related key to the key “Sato” (via steps 1 to 4 in FIG. 9),
"Suzuki" and "Tanaka" are obtained, and from the time when they are set again as the key of interest (step 5), through the same process as described with reference to FIG.
And the related information “document 1”, “document 2”, “document 3” and “document 4” for “Tanaka” (step 6),
The related keys "Noguchi" and "Ando" for the related keys "Suzuki" and "Tanaka" are required (step 7), and the keys "Noguchi" and "Ando" are selected as the indirect related keys for the input key "Sato". (Step 8).

In the above two specific examples, the relation table used in steps 1, 2, 3, and 4 for finding the related key for the input key, and s for finding the related key for the related key
The association tables used in steps 5, 6, 7, 8 are identical. However, their association tables need not be identical. For example, among the specific examples shown here, as the relation table used in steps 5, 6, 7, and 8 in the above specific example, the relation table shown in Table 5 and using a key as a keyword and information as a search expression is shown. By using this, the related keys “Apple”, “Database”, and “Network” related to the input key “Computer” can be obtained not through a document but through a search formula.

FIG. 16 is a functional block diagram of a fourth embodiment of the information search support device of the present invention.

The difference between the functional block and its configuration shown in FIG. 16 and the functional block configuration of the second embodiment shown in FIG. 10 is that the following functional blocks are added.

1) The control unit 8 in the functional block diagram shown in FIG. 14 inputs the related key and the related degree extracted by the related key extracting unit 3 to the related key extracting unit 3 again. And a function unit having related key extraction control means for extracting the degree of relevance.

FIG. 17 is a flowchart showing a procedure of processing performed by the information search support device shown in FIG.

The processing flow in the fourth embodiment shown in FIG. 16 is basically the same as the processing in the second embodiment shown in FIG. 10, and the processing s1, s4, s in FIG.
5, s6, s7, s8, s10, s11, s12 are processes s1, s2, s3, s4, s in the process shown in FIG. 11, which is the process in the second embodiment shown in FIG.
5, s6, s7, s8, and s9. The difference is that the control unit 8 includes processes s2, s3, and s9.

Hereinafter, processing s unique to the fourth embodiment will be described.
2, s3, and s9, and other processes are abbreviated.

Process s1: Same as process s1 in FIG.

Process s2: The input key is set as a key of interest on which a related key is obtained.

Process s3: If the key of interest can be set, the process moves to process s4. If not, the process moves to step s10.

Processing s4 to s8: Processing s2 in FIG.
Same as s6.

Process 9: One of the related keys extracted in process s5 is selected, and is set as a target key from which a related key is obtained.

Processes s10 to s12: Same as processes s7 to s9 in FIG.

In this embodiment, the control unit 8 shown in FIG.
As a result of the presence of the processes s2, s3, and s9 to be performed, processes for finding a related key for a related key and a process for finding the degree of relevance between an input key and a related key are added.
By performing the repetition, it is possible to obtain an effect that more related keys and the degree of association between the input keys and the related keys are transitively obtained. As a result, in the fourth embodiment, the user has related keys with respect to the key designated by the user and ranking information indicating the superiority or inferiority of the related keys,
Furthermore, it is possible to receive transitively determined related keys for the related keys and ranking information indicating the superiority or inferiority of the related keys.

Table 4 for the fourth embodiment shown in FIG.
This will be specifically described with reference to FIG.

Assume that the relation table shown in Table 4 is stored in the relation table storage unit 2 in FIG. 16, and the key (keyword) “computer” is input from the user to the key input unit 1 in FIG. In the information search support device shown in FIG. 16, a related key for the key “computer” is obtained by the related key extracting unit 3. As a result, the related keys directly related to the key "computer" are "Apple", "database", "network", and the related key "banana" for "network" as the indirect related key to the key "computer". Can be

In preparation for calculating the degree of association with the indirectly related key, first, a related key sequence is defined as follows.
In other words, the related key string is a key string obtained as a result of concatenating related keys sequentially extracted based on a certain key in the order of extraction, such as an associated key for a certain key and an associated key for the related key. However, restrict the related key column to not include duplicate keys. Given a set P of related key sequences starting with an existing key a and ending with key a ', the degree of association r (a, a') between keys a and a 'in P is given by the following equation (3). Given by

[0175]

(Equation 3)

[0176] _{However, p: a = a 1,} a '= related key columns to meet the a _n, r (p): Related key column _{p (= [a 1, a} 2, ......,
a _n ]). Using equation (1) described above,

[0177]

(Equation 4) And

In this embodiment, the above equation (3) is used as a relevance calculating means for calculating the inter-key relevance when the relevant key is extended to an indirect relevant key. Equation (3) means that the degree of association between the keys of interest a and a ′ is
Higher according to the number of related key sequences connecting them, and lower according to the length of the related key sequence (ie, the total number of keys that appear).

Here, when the degree of inter-key association between the related keys “Apple”, “Database”, “Network”, and “Banana” and the input key “Calculator” is obtained, “Apple”, “Database”, “ "Network" is in agreement with the above results, with 0.4, 0.4, and 0.2 respectively.
Becomes On the other hand, the degree of association between the input key "computer" and its indirectly related key "banana" is only related to the key sequence from the key "computer" to the key "banana": [computer, network, banana] Therefore, by substituting the degree of association between “computer” and “banana” into equation (3), the degree of association between the input key “calculator” and its indirectly related key “banana” is calculated. 0.13.

In this embodiment, the recommendation order using the inter-key relevance according to the equation (3) is, in this case,
"Apple" and "Database" are the same, then "Network" and "Banana"
The overall rank is determined by the related key rank determination unit 6 in FIG.

The relation key order output unit 7 in FIG.
The ranking information according to the ranking result is output. Here, the output from the related key order output unit 7 is output through an image output device such as a CRT when the output destination is a user who is a human, and is output through a line in another computer system. A method such as transmission as a signal is employed.

FIG. 18 is a functional block diagram of a fifth embodiment of the information search support device of the present invention.

The structure of the functional block shown in FIG.
The difference from the configuration of the functional blocks of the first embodiment shown in FIG. 6 is that the functions and functional blocks described below are added.

1) The combination condition input unit 9 in FIG. 18 is used to extract a related key obtained as a result of combining related key groups for a plurality of input keys input to the key input unit 1 by the user. This is a functional unit having combination condition setting means for inputting a combination condition for designating a combination method.

2) The control unit 8 in FIG. 18 extracts the related key to repeat the processing in the related key extracting unit 3 until a related key satisfying the combination condition input from the combination condition input unit 9 is extracted. A means for sending an input key or a related key for which a related key is requested to the unit 3 is added.

FIG. 19 is a flowchart showing a procedure of processing performed by the information search support device shown in FIG.

The flow of processing performed by the information search support device shown in FIG. 18 is basically the same as the flow of processing in the first embodiment shown in FIG.
s4, s5, and s6 are processes s1, s2, and s in the process shown in FIG. 7 in the first embodiment shown in FIG. 6, respectively.
3 and s4. The difference is that a process s2 performed by the combination condition input unit 9 and a process s3 performed by the control unit 8 in FIG. 18 are provided.

Hereinafter, special processes s2 and s3 in this embodiment will be described.
And the following processing is omitted.

Process s1: Same as process s1 in FIG.

Process s2: A combination condition indicating which related key is output from the related key for each input key is set.

Process s3: If a related key that satisfies the combination condition is extracted, the process moves to process s4. If not, the process moves to step s6.

Processing s4 to s6: Processing s2 to s2 in FIG.
Same as s4.

In this embodiment, s2 performed by the combination condition input unit 9 and processing s performed by the control unit 8 in FIG.
As a result, in addition to the operation of the first embodiment shown in FIG. 6, it becomes possible to specify a related key by a condition in which a plurality of input keys are combined, and it is possible to specify more various related keys by recommendation. .

FIG. 20 is a functional block diagram of a sixth embodiment of the information search support device of the present invention.

1) The relation table designating section 10 in FIG. 20 is a functional section having relation table designating means for inputting relation table designating information for designating a relation table used when a user obtains a relation key.

2) In the control unit 8 in FIG. 20, for each key input to the key input unit 1, the relevant key is input until the relevant key for each relevant table input to the relevant table designation unit 10 is obtained. In order to repeat the processing in the key extracting section 3, a means for sending the related key extracting section 3 an input key or a related key for which a related key is to be obtained and related table designation information for designating a related table to be used is added. ing.

FIG. 21 is a flowchart showing a procedure of processing performed by the information search support device shown in FIG.

The flow of the processing shown here is basically the same as that of FIG.
21 is the same as the processing performed by the first embodiment shown in FIG.
Are common to the processes s1, s2, s3, and s4 in the process shown in FIG. 7, which is the process flow of the first embodiment. The difference is that a process s2 performed by the association table designating unit 10 and a process s3 performed by the control unit 8 in FIG. Less than,
Processes s2 and s3 unique to the present embodiment will be described, and other processes will be abbreviated.

Process s1: Same as process s1 in FIG.

Process s2: Specify related table information for specifying a related table to be used for extracting a related key corresponding to an input key.

Process s3: When the process of retrieving the related keys for all the related tables specified in process s2 is completed, the process moves to process s6. Otherwise, the process moves to step s4.

Processes s4 to s6: Processes s2 to s2 in FIG.
Same as s4.

In the present embodiment, the processing s2 performed by the relation table designating section 10 in FIG. 20 and the processing s3 performed by the control section 8 exist. As a result, in addition to the operation in the first embodiment shown in FIG. It is possible to specify the related table to which the information to be used belongs, and the user specifies what information is required for the related key to be presented as the related key to be presented, and also specifies the related key retrieved in each of those related tables. Comparison becomes possible.

FIG. 22 is a functional block diagram of a seventh embodiment of the information search support device of the present invention.

The difference between the configuration of the functional block of the seventh embodiment shown in FIG. 22 and the configuration of the functional block of the fourth embodiment shown in FIG. 16 is that the following functions are added.

1) Output-related key designation unit 11 in FIG.
Is a functional unit having means for allowing a user to input any of the related key restriction conditions or a combination thereof.

2) The control unit 8 in FIG. 22 is provided with a means for extracting a related key within a range satisfying the relation-restriction condition input to the output related key designation unit 11.

FIG. 23 is a flowchart showing the means of processing performed by the information search support device shown in FIG.

The processing according to the seventh embodiment shown in FIG. 23 is basically the same as the processing in the first embodiment shown in FIG. 6, and the processing s1, s5, s6, and FIG.
s8 is the same as the processing s1, s2, s3, and s4 in FIG. 7 showing the processing in the first embodiment.
The difference is that the process S2 performed by the output-related key designation unit 11 and the processes s3 and s performed by the control unit 8 in FIG.
4 and s7. However, the stop conditions in the processes s2 and s4 in FIG. 23 are, for example, conditions that limit the range of the degree of relevance of the key obtained as the relevant key. Alternatively, the condition may be a condition that specifies a range of the number of repetitions for extracting the related key using the related table, or a condition that specifies a related key that is not presented as the related key. The condition may be configured. When it becomes impossible to further fetch a new related key in a state where these conditions are satisfied, the process for obtaining the related key is stopped.

Hereinafter, processes s2 and s unique to this embodiment will be described.
3, s4, s7, and other processes are abbreviated.

Process s1: Same as process s1 in FIG.

Process s2: A stop condition specified for extracting a related key corresponding to an input key is set. Processing s
3: The input key is set as a key of interest on which a related key is obtained.

Process s4: If the stop condition is not satisfied and it is not impossible to take out a new related key, the process moves to process s5. If not, the process moves to step s8.

Processing s5 to s6: Processing s2 to s2 in FIG.
Same as s3.

Process s7: One of the related keys extracted in process s6 is selected, and is set as a target key from which the related key is obtained.

Process s8: Same as process s4 in FIG.

In the present embodiment, the process s2 performed by the output-related key designation unit 11 and the processes s3, s4, and s7 performed by the control unit 8 in FIG. 23 are present. As a result, the basic embodiment shown in FIG. In addition to the operation of the first embodiment), it is possible to specify a related key to be extracted in more detail.

Next, a specific example of the seventh embodiment described with reference to FIGS. 22 and 23 will be described with reference to Tables 5, 6 and 8 and FIG. Here, the fifth embodiment described with reference to FIGS. 18 and 19, FIGS.
A specific example of the sixth embodiment described with reference to FIG.

Table 5 is a relation table showing the relation between the keyword as the key and the search formula as the related information. As the key, the same key (ie, keyword) as the key in the relation table in Table 4 is used. , And the search expression "banana"
"Banana &Calculator" and "Calculator & Database &Network" (however, the search expression "A &B" indicates a condition that "keys A and B appear together") are registered. The association table shown in Table 5 indicates that, for example, a relationship having an evaluation value 8 as a degree of relevance is established as a relevance relating to a pair of a key “calculator” and a search expression “banana & calculator”. ing. Table 6 is a relation table showing the relation between the keyword as the key and the user as the related information. The key (that is, the keyword) in the relation table of Table 4 is used as the key, and the related information is also displayed. "Sato", "Suzuki", "Tanaka", and "Noguchi" are registered as users. And, in the related table shown in Table 6,
For example, as the relevance of the pair of the key “computer” and the user “Suzuki”, it is shown that the relevance having the evaluation value 3 as the relevance is established.

As the above evaluation values in Tables 5 and 6, various evaluation values such as an evaluation value directly given by a user can be set. Here, the number of uses is used in association with the evaluation value. . In Table 5, the evaluation value 8 is given to the relevancy between the key “calculator” and the search formula “banana & calculator”, which means that the key “computer” has the search formula “banana & calculator”
Indicates that the search has been used eight times in the past. Similarly, in Table 6, the evaluation value 8 is given to the relevancy between the key “calculator” and the user “Suzuki”, which means that the key “computer” has been searched eight times in the past by the user “Suzuki”. Indicates that it is used.

Hereinafter, in order to distinguish the related tables shown in Table 4, Table 5, and Table 6 from each other, they are referred to as “key document related table”, “key search formula related table”, and “key user related table”. Tables.

By preparing in advance a plurality of related tables classified in a plurality of related forms as described above, the tables shown in FIGS. 20 and 21 are designated, and for example, a related key based on the plurality of related tables is set. A form of recommendation becomes possible.

FIG. 24 is a diagram showing an example of an input screen for inputting a key to be recommended by a related key and inputting a condition to be set.

In FIG. 24, there are several input items for setting conditions. These will be described in order from the top in the figure.

In FIG. 24, a first input item is an "input key" item, which is a item for allowing a user to input a key for recommending a related key, and provides a key input section to the user. In FIG. 24, the keyword “calculator” has been input as an input key. This section contains multiple keys, AND, OR,
By inputting a conditional expression linked by NOT or the like, a related key common to a plurality of keys, a related key that is a related key between any of the plurality of keys, or a related key of the key A key or the like which must not be specified can be specified in the form of a conditional expression, and the fifth embodiment (FIGS. 18 and 19) is provided to the user.
) Is provided.

The "relation table" item, which is the second input item, is used to input a relation table name designating a relation table to be used for obtaining a relation key. The relation table in the sixth embodiment shown in FIG. A table designating unit 10 is provided. FIG. 24 shows a state in which all of the category names of the related information appearing in the association table, namely, “document”, “search formula”, and “user” are input. For each related table (“key document related table”, “key search formula related table”, and “key user related table”), a condition is set that recommends a related key. As another specification method, a method of recommending a related key obtained as a result of applying a plurality of specified related tables to a related key sequentially according to a specified order, or from a plurality of specified related tables. Common related keys required,
Alternatively, various conditions can be set, such as a method of recommending a related key required from any one of the related tables, a method of recommending a related key excluding a related key extracted from a specific related table.

The third input term “relevance range” is:
This is a section for inputting a value for specifying a range of the degree of relevance of the relevant key to be recommended, and provides an example of the output-related key specifying unit 11 in the seventh embodiment shown in FIG. FIG. 24 shows a state in which a range in which the degree of association is 0 or more and 0.6 or less is specified. Here, it is also possible to specify a range obtained by combining a plurality of ranges.

The fourth input item “recommended number” is an item for inputting values for designating the lower limit and the upper limit of the number of related keys to be recommended, and outputs the values in the seventh embodiment shown in FIG. Another example of the related key specifying unit 11 is provided. FIG. 24 shows a state in which it is specified to recommend one or more related keys in a range of 10 or less. This condition setting can also be specified in a range obtained by combining a plurality of ranges, as described above.

The fifth input item, the “number of repetitions” item, is a item for inputting a value specifying the range of the number of times the relation table can be repeatedly applied to obtain the related key to be recommended. An example of the output-related key designation unit 11 in the seventh embodiment shown in FIG. 22 is provided. FIG. 24 shows a range in which the number of repetitions is one or more and one or less, that is, a state in which it is specified that the number of repetitions is limited to one without repetition. Instead of directly specifying the number of repetitions as shown here, various specifying methods such as setting a default value or specifying the previous number of times are possible.

[0230] The last entry term, "unrelated key" term, is
This is a section for inputting a key name for specifying a key not to be used as a related key. This also provides an example of the output related key specifying unit 11 in the embodiment shown in FIG. FIG.
In FIG. 4, the key name "Apple" has been input, which means that the condition that the keyword "Apple" is not recommended as a related key has been set. In this condition setting, instead of directly specifying an unrelated key as shown here, it is also possible to indirectly specify the degree of relevance for selecting an unrelated key, a relation table, and the like.

In the case of the device form shown in FIG. 3 or FIG. 5, these input means are realized using an input device 103 such as a keyboard, and the means for presenting an input screen is realized using an output device 104 such as a CRT. In the case of the device form shown in FIG. 4, the receiving device 106 that receives information transmitted from the search device 400_1 or the like serves as an input unit, and displays the screen of the transmission device 107 as shown in FIG.
The function of transmitting to 1 etc. plays the role of output means. The actual input operation and screen output are performed by the search devices 400_1,.
_N.

In the embodiment shown in FIG. 22, the control unit 8 which has received a key for which a related key is to be recommended from a user and a constraint condition for determining the related key is used by the control unit 8 for the key received from the user. The related key extraction unit 3 and the association degree calculation unit 5 are operated within a range in which the key satisfies the constraint condition.

Table 9 shows the degree of association between the key “computer” calculated using the above equation (1) and the associated keys “banana”, “database”, and “network” in each of the above three association tables. 5 is a table showing values of the. However, the key "Apple" is excluded from the table because it is designated as an unrelated key. For example, the degree of association between the key “computer” and the associated key “database” when the key search expression association table is used is a value 0.56 shown in the center of Table 9.

[0234]

[Table 9]

FIG. 25 is a diagram showing an example of a display screen to the output device of the processing result for each related table corresponding to Table 9.

The processing for obtaining the relevant keys by the relevant key extracting unit, the processing performed by the relevant degree calculating unit for obtaining the relevant degree, and the processing for controlling them by the control unit are performed in any of the device forms shown in FIGS. Is also realized by the CPU 101, the memory 102, and the program stored in the memory 102.

Specifically, when the results shown in Table 9 are obtained, the relation between the key “computer” and the related keys “banana”, “database”, and “network” in each relation table is determined. , In terms of key document association tables, "banana", "network", "database"
In the order of priority. Similarly, from the viewpoint of the key search expression relation table, "database" and "network" are higher than "banana", and "database" and "network"
Are tied. Also, from the viewpoint of the key user relation table, the priority is higher in the order of “banana”, “database”, and “network”. The order between related keys is determined as described above.

FIG. 26 is a functional block diagram of an information search support device according to an eighth embodiment of the present invention.

The difference between the configuration of the functional block of the eighth embodiment shown in FIG. 26 and the configuration of the functional block of the second embodiment shown in FIG. 10 is that the following functions are added.

1) The update information input unit 12 in FIG. 26 is a functional unit having an association table update information input unit for inputting the update content of the evaluation value of the association table.

2) The related table updating unit 13 in FIG. 27 is a functional unit having related table updating means for updating the related table stored in the related table storage unit 2 in accordance with the updated contents of the evaluation value of the related table. is there.

Hereinafter, the update information input section 12 and the related table update section 13 will be described in detail.

The update information input unit 12 is realized by the input device 103 in the case of the device configuration shown in FIGS.

In the case of the device configuration shown in FIG. 4, the input devices are provided in the search devices 400_1,..., 400_n, and the input information from the input devices is transmitted to the information search support device 100B via the communication path 300. Received by the receiving device 106. Therefore, the receiving device 106 of the information search support device 100B corresponds to the update information input unit 12.

Here, description will be made assuming that the apparatus has the configuration shown in FIG. The related table update unit 13 receives the related table update information input to the update information input unit 12, and updates the contents of the related table stored in the related table storage unit 2 (the database 104 illustrated in FIG. 3). .

FIG. 27 is a diagram showing a presentation example of the update input screen.

The contents of the update shown in FIG. 27 are the same as those in the related table in which the related information is a document, that is, in the key document related table.
This indicates that the degree of association between key information between the key “computer” and the information “document 2” is updated to “3”.

In the present embodiment, by having such an association table updating means, it is possible to directly express the evaluation of the usefulness and the association of the key with respect to the associated information by the user. Then, a key recommendation based on the evaluation by such a user can be received.

Also, instead of the user directly updating the association table as described above, keys, information, and the degree of association between key information are extracted from the history of information collection performed by the users, and the results are extracted according to the result. Alternatively, the device may automatically update the association table. For example, according to the evaluation criterion that “a keyword k appears with high frequency in a search expression including a document d in a search result means that the keyword k and the document d are highly related”, The degree of association between key information, that is, the evaluation value of the document d with respect to the keyword k is calculated by using the number of times n of the search formula including the keyword k given by the user to the search engine such that the document d is included in the search result
Give in. This change in the number of uses n can be automatically extracted by continuously monitoring the pair of the search formula given by the user to the search engine and the search result returned to the user by the search engine. . Therefore, the update information input unit 12 in FIG. 26 includes means for automatically extracting a change in the number of uses n and automatically setting conditions for updating the related table in accordance with the change. Updates are possible. As a result, the user does not need to perform the evaluation work and the update work for each pair of the key and the information, and can receive the key recommendation. It should be noted that if the number of uses is directly written in the related table, the number written in the related table only increases, so that the sum of all numerical values in one related table is 1.0 in the related table, for example. The number of times of use may be standardized and written.

As another method for updating the association table,
An intermediate method between the method described with reference to FIG. 27 and the method in which the user directly writes in the related table and the method in which the user automatically extracts and updates automatically from the information collection history described above may be adopted.

For example, a key or key group to be input at the time of an actual search may be added with information of importance, and the related table may be updated based on the information of importance.
For example, a search expression "banana 10 & computer 20" obtained by adding importance information to a search expression "banana &computer" including a key "banana" and a key "calculator" is input at the time of actual search. '10', '20' shown there
Based on the numerical value indicating the degree of importance, the evaluation value of the key search formula association table as shown in Table 5 may be updated little by little.

Alternatively, the user does not add importance information to the search formula, but automatically assigns importance according to the order added to the search formula, and automatically updates the related table according to the assigned importance. You may. For example, "Calculator &
If the search expression "database &network" is input, the importance is '10' for the computer which is the first keyword in the search expression, the importance is '5' for the database which is the second keyword, and the third is The importance can be automatically assigned to the keyword “network” according to the order in which it appears in the search formula, such as importance “2”.

Furthermore, when a certain document is found as a result of performing a search by inputting a certain search expression, the importance of the found document is entered. Based on the entered importance, Table 4 is entered. May be updated.

FIG. 28 is a diagram showing an embodiment of a storage medium of the information search support program of the present invention.

An information search support program 700 is stored in the information search support program storage medium 600 shown here. This information search support program 700 is shown in FIG.
5 is a program for operating an apparatus configured as a computer system shown in FIG. 5 as an information search support apparatus for realizing the present invention. In the example shown in FIG.
It is composed of a relation table creating means 701, a related information extracting means 712, a related key extracting means 703, and others. The related table creating means 701 is a program having a function of creating and updating the various related tables described above, and the related information extracting means 702 receives an input key, refers to the related table, and inputs the input key. This is a program for extracting relevant information related to a key.
This is a program for extracting a related key, which is a key different from the input key, which is further related to the related information extracted by the related information extracting means 702. In addition to the above, the information search support program 700 shown in FIG. 28 causes a computer system having the device configuration shown in FIGS. 3 to 5 to operate as the information search support device of the first to eighth embodiments described above. It consists of all necessary programs.

The information search support program storage medium 600 shown in FIG. 28 is, for example, a portable storage medium 200 such as an optical disk or a magneto-optical disk shown in FIGS.
In addition, when the information search support program 700 stored in the portable storage medium 200 is uploaded to the memory 102 shown in FIGS.
2 also corresponds to the information search support program storage medium according to the present invention.

[0257]

As described above, according to the present invention,
At the next point in time, a key that takes into account the variety of user intentions can be recommended as a key to be used for the search. An object of the present invention is to find a related key for a key such as a keyword specified by a user and recommend it as a key to be used for a search at the next point in time. If a related key exists under a specific related table or a combination of multiple related tables, conversely, a plurality of keys specified by the user will be displayed under the specific related table or combination of related tables. Can be viewed as related via a common related key, which means that the present invention also provides a way to discover the relationship between keys.

Further, in the present invention, if an association table with an evaluation value is adopted, the user can set the degree of association between key information as an evaluation value of the usefulness of information with respect to the intention represented by the key. It is possible to prepare a related table in which evaluation values reflecting the evaluation based on the evaluation are set. As a result, based on the correspondence between the key represented by the related table and the related information, the relationship between the keys is evaluated, and based on the evaluation result,
By determining a key to be recommended for a key input by a user, a key recommendation that more accurately reflects the user's intention can be performed.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a related key.

FIG. 2 is an explanatory diagram of a basic principle for extracting a related key according to the present invention.

FIG. 3 is a diagram showing one embodiment of the device configuration of the information search support device of the present invention.

FIG. 4 is a diagram showing another device configuration of the information search support device of the present invention.

FIG. 5 is a diagram showing another device configuration of the information search support device of the present invention.

FIG. 6 is a functional block diagram of a first embodiment of the information search support device of the present invention.

FIG. 7 is a flowchart illustrating a procedure of a process performed by the information search support device illustrated in FIG. 6;

FIG. 8 is an explanatory diagram showing a specific example of the first embodiment.

FIG. 9 is an explanatory diagram showing another specific example of the first embodiment.

FIG. 10 is a functional block diagram of a second embodiment of the information search support device of the present invention.

11 is a flowchart illustrating a procedure of a process performed by the information search support device illustrated in FIG. 10;

FIG. 12 is a functional block diagram of a third embodiment of the information search support device of the present invention.

FIG. 13 is a flowchart showing a procedure of a process performed by the information search support device shown in FIG.

FIG. 14 is a schematic diagram showing a latter half of a specific example of a processing flow in the third embodiment.

FIG. 15 is a schematic diagram showing a latter half of another specific example of the processing flow in the third embodiment.

FIG. 16 is a functional block diagram of a fourth embodiment of the information search support device of the present invention.

FIG. 17 is a flowchart illustrating a procedure of a process performed by the information search support device illustrated in FIG. 16;

FIG. 18 is a functional block diagram of a fifth embodiment of the information search support device of the present invention.

FIG. 19 is a flowchart illustrating a procedure of a process performed by the information search support device illustrated in FIG. 18;

FIG. 20 is a functional block diagram of a sixth embodiment of the information search support device of the present invention.

FIG. 21 is a flowchart illustrating a procedure of a process performed by the information search support device illustrated in FIG. 20;

FIG. 22 is a functional block diagram of a seventh embodiment of the information search support device of the present invention.

FIG. 23 is a flowchart showing a procedure of processing performed by the information search support device shown in FIG.

FIG. 24 is a diagram illustrating an example of an input screen for inputting a key to be recommended for a related key and inputting a condition to be set;

FIG. 25 is a diagram illustrating an example of a display screen to the output device of a processing result for each association table corresponding to Table 9.

FIG. 26 is a functional block diagram of an information search support device according to an eighth embodiment of the present invention.

FIG. 27 is a diagram illustrating a presentation example of an update input screen.

FIG. 28 is a diagram showing an embodiment of a storage medium of the information search support program of the present invention.

[Explanation of symbols]

REFERENCE SIGNS LIST 1 key input unit 2 related table storage unit 3 related key extraction unit 4 related key output unit 5 relevance calculation unit 6 related key order determination unit 7 related key order output unit 8 control unit 9 combination condition input unit 10 related table designation unit 11 Output related key designation unit 12 Update information input unit 13 Related table update unit 100A, 100B, 100C, 100D,..., 100
N information search support device 101 CPU 102 memory 103 input device 104 output device 105 database 106 reception device 107 transmission device 200 portable storage medium 300 communication path 400_1, ..., 400_n search device 500 database 600 information search support program storage medium 700 information search Supporting program 701 Related table creating means 702 Related information extracting means 703 Related key extracting means

Claims (14)

[Claims] 1. An association table storing an association table in which information on presence / absence of association between each key constituting a predetermined key group and each related information constituting a predetermined related information group is stored. A storage section, a key input section for inputting a key, receiving a key input from the key input section, referring to a relation table stored in the relation table storage section, and relating to the input key being the input key. A related key extracting unit that extracts a related key that is different from the input key, and a related key output unit that outputs the related key extracted by the related key extracting unit. Characteristic information search support device. 2. The information search support device is connected to a communication channel, wherein the key input unit is a receiving unit that receives a key transmitted via the communication channel, The key output unit is a transmission unit that transmits the related key extracted by the related key extraction unit to the communication path toward a transmission source that transmits the input key received by the key input unit, The information search support device according to claim 1, wherein 3. The association table storage unit stores an association table in which not only the presence or absence of association between each key and each piece of related information but also the degree of association between key information that further indicates the degree of association. The information retrieval support device further includes a reference table stored in the relation table storage unit, wherein the input key is extracted by the relation key extraction unit in relation to the input key. A relevance calculating unit for obtaining a relevance between keys indicating a degree of relevance between the input keys and a relevance calculating unit for obtaining a relevance between the input keys. A related key order determining unit that determines the order of relevance to the input key among the plurality of related keys according to the inter-key related degree, and the order of the related keys determined by the related key order determining unit. Related key to output information Information retrieval support apparatus according to claim 1, characterized in that a position output section. 4. The related key extracting unit extracts a related key other than the input key and the already extracted related key, further related to related information further related to the related key extracted by the related key extracting unit. The information search support device according to any one of claims 1 to 3, further comprising a control unit for causing the information search to be performed. 5. A combination condition input unit for inputting a plurality of keys, wherein the information search support device further includes a combination condition input unit for inputting a combination condition for the plurality of input keys input from the key input unit. And a controller for causing the related key extracting unit to extract a related key that matches the plurality of input keys input from the key input unit and the combination condition input from the combination condition input unit. Any one of claims 1 to 4
Information search support device described in the section. 6. The association table storage unit stores a plurality of types of association tables, and the information search support device further includes one or more association tables for specifying one or more association tables to be searched. A table designating unit, and the related key extracting unit, wherein a related key of one or more related tables designated by the related table designating unit among a plurality of types of related tables stored in the related table storage unit. A control unit for performing extraction is provided.
The information retrieval support device according to any one of the above. 7. An output-related key specifying unit for specifying a relevant key restriction condition for restricting a relevant key extracted by the relevant key extracting unit; The information retrieval support device according to any one of claims 1 to 4, further comprising a control unit configured to extract a related key in a range satisfying the related key restriction condition. 8. An update information input section for inputting update information for updating a relation table stored in the relation table storage section, and the relation table storage section based on update information input from the update information input section. A relation table updating unit for updating a relation table stored in the storage unit.
The information retrieval support device according to any one of the above. 9. The information search support device according to claim 8, wherein the update information input unit includes a unit that generates the update information based on a past information collection history. 10. The related key extracting unit extracts related information related to the input key by referring to a predetermined related table,
The information according to any one of claims 1 to 9, wherein a related key further related to the extracted related information is extracted by referring to the same related table as the predetermined related table. Search support device. 11. The relation table storage unit stores a predetermined first relation table, a key group having a genre different from a key group of the first relation table, and the first relation table. And a second related table including a related information group of the same genre and a related information group of the same genre, wherein the related key extracting unit stores related information related to the input key in the first related table. 10. The method according to claim 1, wherein the related key is extracted by referring to a related table, and a related key further related to the extracted related information is extracted by referring to the second related table. Information search support device described in the section. 12. An association table in which an association table storing information on the presence or absence of association between each key constituting a predetermined key group and each associated information constituting a predetermined related information group is stored. A key input unit for inputting a key, which is connected to a device having a storage unit via a communication path, receives a key input from the key input unit, and refers to a relation table stored in the relation table storage unit A related key extracting unit that extracts a related key that is a key different from the input key, further related to related information related to the input key that is the input key, and a related key extracted by the related key extracting unit. An information retrieval support device comprising a related key output unit for outputting a key. 13. An input key serving as a search trigger is received, and information on the presence or absence of a relationship between each key constituting a predetermined key group and each related information constituting a predetermined related information group is described. Related information extracting means for referring to the obtained related table and extracting related information related to the input key; and a key different from the input key, further related to the related information extracted by the related information extracting means. An information retrieval support program storage medium storing an information retrieval support program having a related key extracting means for extracting a related key. 14. The information search support program storage medium according to claim 13, wherein said information search support program has a relation table creation means for forming said relation table.