JP2001142889A - Device and method for mediating information - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an information mediating device, which can appropriately mediate information even when an object for which information is mediated is a huge quantity. SOLUTION: A profile received by a profile input part 101 is stored in a profile group managing part 102, and plural profiles are successively extracted from the stored profile group and collated by a profile collating part 103. The set of profiles collated by the profile collating part 103 is stored in a collated result holding part 104 and while referring to a thesaurus knowledge managed by a thesaurus managing part 106, only the information of high relation is selected out of the set of collated profiles by an information mediating part 105. The collated result selected by the information mediating part 105 is transmitted to the relevant user by a mediated information output part 107.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information intermediary apparatus and an information intermediation method for collating preferences and preferences between users and intermediating conceptually similar items using a thesaurus.

[0002]

2. Description of the Related Art Conventionally, the most commonly used technique for acquiring information reflecting user preferences and preferences on the Internet or the like has been an information search technique. But,
With multi-channel web pages and digital broadcasting,
Search results have become enormous, and the burden on users to reach information that suits their tastes is ever-increasing. In order to improve such a situation, a technique for automatically changing a search key according to the amount of search results (Japanese Patent Laid-Open No. Hei 8
-6970) has been proposed.

On the other hand, there is a technique called information mediation as an approach different from information retrieval. In this method, personal preferences are stored in the system as information in advance, and based on this information, information is collated with information from another user or another information source, and the information is mediated.
For example, a system called FireFly (U.S. Pat. No. 5,749,081) creates pairs of user preferences, adds their relevance as weights, and introduces the preferences. A search is made to find one whose preference appears in one of the pairs, and to have the other party be ranked and introduced according to the weight ratio.

[0004]

However, in the former of the prior art, since the improvement is only within the framework of "information search", the "requirement of information required by the user" which search technology essentially has Conditions must be clarified every time. "

[0005] Further, in the latter case of the prior art, "information mediation" is used.
When the number of objects to be performed becomes enormous, it is not always possible to perform appropriate mediation because it is not possible to use the perspective relationship of concepts using semantic knowledge. Furthermore, since this technology is based on collation of a target character string, effective mediation can be achieved by actively using the target attribute information and the semantic perspective relation of the attribute information. Not in.

An object of the present invention is to solve the above-mentioned problems, and an object of the present invention is to provide an information intermediation device capable of performing appropriate intermediation even when the number of objects to be intermediated is enormous.

[0007]

In order to solve this problem, according to the present invention, information is provided based on an information mediation technique, not an information search technique.

[0008] In order to perform information mediation, a group of items, which is a user's favorite, called a profile is held, and a plurality of "attribute information" is given to each item. By using this attribute information and thesaurus knowledge (conceptual perspective relationship) positively, effective mediation taking into account semantic perspective relationship is performed.

Mainly, a profile input via a network is collated with another profile, attribute information of the collated profile is extracted from the collated profile items, and the similarity of the concept of the attribute is determined by a thesaurus. The management unit makes a decision, and mediates only those with a high similarity.

[0010] Conversely, it is also possible to intentionally select and mediate items having conflicting attribute information. further,
In addition to the profile input by the user, a profile is created from content attribute information provided by the content provider, and collation and mediation can be performed in the same manner as the profile input by the user.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An information mediating apparatus according to the present invention according to the first aspect of the present invention provides a profile input unit for receiving a profile, and a profile group for receiving and managing the profile by using the profile received by the profile input unit as an input. A management unit, a profile matching unit that sequentially retrieves a plurality of profiles from the profile group stored in the profile group management unit and performs matching, and a matching unit that receives information on a set of profiles matched by the profile matching unit and stores the result. A result group holding unit, an information mediating unit that receives the matching result stored in the matching result group holding unit, selects only highly relevant information from the matched profiles while referring to the thesaurus knowledge, and an information mediating unit. And a mediation information output unit for transmitting the mediation information output from the user to the user. According to this configuration, in order to perform information mediation, a set of items of a user's favorite called a profile is held, the input profile is compared with another profile, and the profile is checked from among the items of the profile that has been compared. Then, the attribute information of the item is extracted, the similarity of the concept of the attribute is determined by the thesaurus management unit, and only those having a high similarity are mediated. Therefore, effective mediation in consideration of the semantic perspective relationship can be performed.

According to a second aspect of the present invention, there is provided the information intermediation device according to the first aspect, wherein the information intermediation unit has a configuration for selecting only information having an opposite relationship. ing. According to this configuration, it is possible to intentionally select and mediate items having conflicting attribute information. This is effective, for example, when you want to watch a movie that you are tired of and thrilled by a thrilling movie.

[0013] According to a third aspect of the present invention, there is provided an information intermediation apparatus according to the first aspect, further comprising: a content attribute information input unit for receiving content attribute information;
A content attribute information group management unit that holds and manages content attribute information by inputting the content attribute information received by the content attribute information input unit, a content attribute information classification information input unit that inputs classification information of the content attribute information, It has a configuration in which the attribute information classification information and the information managed by the content attribute information group management unit are input and a content attribute information classification unit that classifies the content attribute information is provided. According to this configuration, it is possible to create a profile from not only the profile input by the user but also the content attribute information provided by the content provider, and perform collation and mediation in the same manner as the profile input by the user.

[0014] According to a fourth aspect of the present invention, there is provided an information intermediation method according to the present invention, wherein: a process of setting an input profile as a first variable; When an untested profile exists, extracting one of the unexamined profiles and setting it to a second variable; and collating the first and second variables. Recording and storing copies of the profiles of the first and second variables, returning the second profile to the profile group management unit when there is no collation item, and when a collation profile pair exists Selecting only highly relevant information from the collated profiles while referring to the thesaurus knowledge. It is. According to this configuration, the input profile is compared with another profile, the attribute information of the item is extracted from the matched profile items, and the similarity of the concept of the attribute is determined by the thesaurus management unit. , And mediates only those with high similarity. Therefore, effective mediation in consideration of the semantic perspective relationship can be performed.

According to a fifth aspect of the present invention, in the information intermediation method according to the fourth aspect, the step of selecting only highly relevant information from the collated profiles includes the step of selecting the collated item. Checking whether there is an untested item in it, checking whether the genre attribute of the item exists in the matching genre table, and if the genre attribute does not exist in the matching genre table, It has a configuration including a step of adding a new genre to the collation genre table and a step of selecting mediation information when there is no untested item in the collated items. According to this configuration, effective mediation can be performed for all untested items and genre attributes in consideration of the semantic perspective relationship.

[0016] According to a sixth aspect of the present invention, in the information intermediation method according to the fifth aspect, the step of selecting the intermediation information includes the step of selecting the non-collated items of the second profile. A step of checking whether or not an unchecked item exists; a step of setting the corresponding item as a third variable if any, and a genre attribute of the corresponding item as a fourth variable; Checking if the item exists;
Setting a genre attribute of the corresponding item as a fifth variable if there is an untested item; checking whether the fourth and fifth variables are the same concept; if not, the fourth and fifth variables on the thesaurus; Checking whether the fifth variable is a sibling concept or a descendant concept, and if it is a sibling concept, confirming that it is not an opposing sibling concept; and if the fourth and fifth variables are the same concept, or If the concept is a sibling concept having a similar relationship, or if the concept is a descendant concept, a step of selecting a third variable as mediation information is provided. According to this configuration, it is possible to effectively mediate items having attribute information belonging to the same genre and attribute information belonging to sibling concepts or descendant concepts having a similar relationship.

[0017] According to a seventh aspect of the present invention, in the information intermediation method according to the fifth aspect, the step of selecting the intermediation information includes the step of selecting a non-collated item of the second profile. A step of checking whether or not an unchecked item exists; a step of setting the corresponding item as a third variable if any, and a genre attribute of the corresponding item as a fourth variable; Checking if the item exists;
Setting a genre attribute of the item as a fifth variable when there is an untested item; determining whether the fourth and fifth variables have ancestral concepts that conflict with each other; Selecting a third variable as intermediary information when there is a conflicting ancestor concept. According to this configuration, it is possible to intentionally select and mediate items having conflicting attribute information. The information mediating method according to the present invention described in claim 8 is:
5. The information intermediation method according to claim 4, further comprising a step of inputting a plurality of content attribute information and a step of classifying the plurality of content attribute information with a predetermined attribute, wherein the classified content attribute information is input. It has a configuration for processing as a profile. According to this configuration,
A profile can be created from not only the profile input by the user but also content attribute information provided by the content provider, and collation and mediation can be performed in the same manner as the profile input by the user.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

(Embodiment 1) A first embodiment of the present invention will be described with reference to FIG. FIG. 1 is a conceptual diagram illustrating a configuration of an information mediating apparatus according to the embodiment.

The apparatus is stored in a profile input unit 101 for receiving a profile (information such as user preference), a profile group management unit 102 to which the profile received by the profile input unit 101 is input, and a profile group management unit 102. A profile collating unit 103 and a profile collating unit 10 for sequentially extracting a plurality of profiles, for example, two profiles as one set from the profile group, and performing collation.
3. The information of the set of profiles actually collated in step 3 is input and the collation result group holding unit 104 that holds the result is input. The collation result stored in the collation result group holding unit 104 is input and managed by the thesaurus management unit 106 described later. An information intermediation unit 105 that performs a process of selecting only highly relevant information from a set of collated profiles while referring to the thesaurus knowledge to be compared, and an information intermediation unit 105 that manages thesaurus knowledge and
Thesaurus management unit 106 that provides thesaurus knowledge in response to a request from the user, and the mediation information output unit 1 that transmits the collation result adjusted by the information mediation unit 105 to the user and displays the result.
07.

The input to the profile input unit 101 and the output from the mediation information output unit 107 are mainly performed through a network, and the user performs input and output in a web or mail format. May be used. Further, the profile input unit 101 is used not only to input a profile but also to change or delete a profile already stored in the profile group management unit 102.

The structure of information handled by the information intermediary device configured as described above will be described below. FIG.
Is an example of a profile input from the profile input unit 101 and stored in the profile group management unit 102. In this embodiment, a profile is information in which a plurality of item groups preferred by a certain user are put together. In FIG. 2, the user A first has an item (row) whose ID is 1;
The name attribute of this item is Blade Runner. This name attribute represents the name of the item that each user prefers. The genre attribute of this item is SF, and the media attribute of this item is movie. That is, the information that the user A prefers the SF movie named Blade Runner is described in the first item of the profile of the user A. The same applies to the second and subsequent items. A collection of all items of the user A in this manner is called a profile of the user A. These profiles are input to the profile input unit 101 and stored and managed in the profile group management unit 102. That is, the profile group management unit 102 stores a large number of profiles.

Next, the operation of the present apparatus will be described with reference to the processing flow of FIG. 3 and the configuration diagram of FIG. Profile input unit 1
If there is a new input in 01, it will trigger
The flow of FIG. 3 is started. First, the newly input profile is set in the variable A of the profile matching unit 103 (step 301), and then it is checked whether the profile group management unit 102 has an untested profile (step 302). In this example, it is assumed that the profile of the user A in FIG. 2 is newly input.

In step 302, if there is an untested profile, one is taken out and set in the variable B of the profile matching unit 103 (step 30).
3). Next, the profile matching unit 103 performs a matching process on the variables A and B (step 304), and checks whether there is a matching item (item having the same content) (step 305). In this embodiment, the matching of the collation items is determined by using IDs assigned uniquely to all items handled by the apparatus. In addition, a method of discriminating by matching character strings is also possible.

When the collation items exist, the collation items are recorded, a copy of the profile of the variables A and B is made, and the copies are stored in the collation result group holding unit 104 (step 30).
6). On the other hand, when there is no collation item, the profile B is returned to the profile group management unit 102 (step 3).
07). Regardless of whether the process proceeds to step 306 or step 307, the process returns to step 302, and the process is repeated until there is no untested profile.

In step 302, if there is no untested profile, the variable A is stored in the profile group management unit 102 (step 308), and all the profiles in the profile group management unit 102 are set to the untested state (step 302). 309) to return to the initial state.

When a series of collation tests are completed, it is checked whether or not a collation profile pair exists in the collation result group holding unit 104 (step 310). Performs the mediation processing of the information, and outputs the result from the mediation information output unit 107 (step 311).

The information mediation processing performed in step 311 will be described later.

Referring to FIG. 4, in step 304 of FIG.
The example of the result of having performed the collation processing of the variable A and the variable B is shown. The table on the left is the profile of the user A and uses the same information as in FIG. 2 (however, the description is symmetrical so that the collation part is easy to understand). The table on the right shows the profile of the user B, which is set in step 303 of FIG. Items of ID = 1, ID = 2, ID = 3 are collated. A set of two profiles having collation items is copied and stored in the collation result group holding unit 104 in FIG. 1 together with information on which ID is collated. The PID in FIG. 4 means a profile ID, which is uniquely maintained in the profile group management unit 102 of the information mediation device.

In FIG. 4, since profile A and profile B have three collation items, there is a high possibility that their interests and preferences are similar to each other. Therefore, there is a possibility that interests and preferences of other items that have not been collated are similar.
However, there is no possibility that it is completely different. In the present invention, in order to solve the above-described problem, the mediation of information is efficiently performed by using a thesaurus and considering conceptual perspective. This method will be described with reference to FIG.

FIG. 5 shows a part of the thesaurus used in this embodiment. It shows an example of a concept tree relating to genre attributes of works having a story, such as movies, TV dramas, and books. A group of concepts that exists immediately below a certain concept node is called a “sibling concept”. The relationship between "brother concepts" includes "brother concepts in conflict" and "similar brother concepts"
There are two types: In FIG. 5, the nodes directly below the white circles represent concepts that conflict with each other, and the nodes directly below the black circles represent sibling concepts that are similar to each other. For example, thrill suspense is an at-home "contradictory sibling concept" and hard-boiled and inferring is "similar sibling concept".

For a certain concept node P, a concept node Q that exists below the concept node P is called a “descendant concept” of P. For example, psychohorror is both a horror's “offspring concept” and a thrilling suspense's “offspring concept”.

An information mediating method using a thesaurus will be described with reference to the tables shown in FIG. The processing flow of FIG. 6 is a detail of step 311 of the processing flow of FIG. In this embodiment, the description will be made on the premise that there is one set of “matched profiles”. However, there are naturally cases where there are two or more sets of “matched profiles”. In that case, the processing flow of FIG. This is repeated until there is no “matched profile”.

First, a "collation genre table" for storing the genre attribute of the collated item is initialized (step 601). FIG. 7 shows an example of the “collation genre table”. It is empty immediately after the initialization, and the contents are filled as a result of the processing flow of FIG. In this example, SF, horror, and hard boiled are registered in the “collation genre table” as the “genre attribute” of the collated item.

Returning to the processing flow of FIG. 6, it is checked whether or not an unchecked item exists among the checked items next (step 602). If so, the genre attribute of the item is set to “matching genre”. It is checked whether it already exists in the "table" (step 603). If there is an untested item, the test result is determined (step 60).
4) If the genre attribute does not yet exist in the "collation genre table", it is added to the "collation genre table" as a new collation genre (step 60).
5) Return to step 602. If the genre attribute determined in step 604 exists in the “collation genre table”, the process returns to step 602 without doing anything since it has already been registered in the “collation genre table”. Thus, Steps 602 to 605 are repeated, and Step 60
If there are no more untested items in step 2, step 6
Proceeding to 06, a process of selecting mediation information is performed. This will be described with reference to the processing flow of FIG.

The processing flow of FIG. 8 is a detailed version of the processing flow step 606 of FIG.

FIG. 8 assumes a case where the profile information of the user B is relayed to the user A. First, it is checked whether or not an unchecked item exists in the items that have not been checked (non-checked items) of the profile B (step 8).
01), if present, the information mediation unit 105 in FIG.
The above item is set in the variable XX existing inside the
At the same time, the genre attribute of the corresponding item is set in a variable X (step 802). Next, the above-mentioned "collation genre table" is checked to check whether there is any untested item (step 803). If there is an untested item, the genre attribute of the item is set to the variable Y existing inside the information mediating unit 105 in FIG. 1 (step 804).

Next, the process proceeds to step 805, where the information mediating unit 105 checks whether X and Y are the same concept. If not, the information mediating unit 105 refers to the thesaurus management unit 106 and determines whether X and Y are "the same concept" on the thesaurus. It tests whether it is a "brother concept" or X is a "descendant concept" of Y. Here, if X and Y are “brother concepts”, it is confirmed that they are not “opposite sibling concepts”. That is, if X is the same concept as Y,
Or, if X is a “similar concept of siblings” to Y, or if X is a “descendant concept” of Y, then XX
Is determined as the transfer information to the user A, and the transfer information output unit 10
7 (step 806).

For example, the 2001 space journey with ID = 10 has a genre attribute of SF, and SF exists in the collation genre table. Therefore, it is recorded as mediation information. The Pelican document with ID = 13 has a genre attribute of inference, and inference does not exist in the collation genre table. However, by referring to the thesaurus, SF,
Horror, hard-boiled "similar brother concept"
It turns out that it is. Therefore, it is recorded as mediation information.

The smile of ice with ID = 14 has a genre attribute of psycho horror, and psycho horror does not exist in the collation genre table. However, by referring to the thesaurus, it can be said that it is a horror "child descendant concept". I understand. Therefore, it is recorded as mediation information. In this way, the selection of the mediation information is performed.

If Y does not satisfy the condition in step 805, the process returns to step 803, and the same processing is repeated for the next unexamined item in the collation genre table.

When all the items in the "collation genre table" have been inspected in step 803, all the items in the "collation genre table" are returned to unexamined (step 80).
7) Return to step 801 to repeat the process for the next non-collation item.

When it is determined in step 801 that all non-collation items should be relayed, it is checked in the relay information output unit 107 whether relay information exists (step 808). If so, the mediation information output unit 107 outputs to the user A (step 809). If there is no mediation information in step 808, nothing is performed. That is, in this case, the mediation processing is not performed.

The method of mediating information while referring to the thesaurus knowledge using the processing flow of FIG.
The example of mediating to the user A based on the profile described above has been described. The same processing is performed in the opposite case. In this information intermediary device, information intermediation is treated as mutual processing for collated profiles.

FIG. 9 shows a conceptual diagram of the mediation of information for collation.
The two circles represent the profile A and the profile B, respectively, and the overlapping portion of the circles corresponds to the matching portion. The elements in the circle represent the item ID and the genre attribute.
The rectangular area “identical collation genre” is an item that is subject to information mediation because it belongs to exactly the same genre.
The rectangular area called “neighborhood collation genre” is an item that is a target of information mediation because it is in the neighborhood (“similar sibling concept” or “descendant concept”) with reference to the thesaurus. The distant verification genre will be described in a second embodiment.

Further, if the user A prefers only a movie as a medium by using the attribute of the item called a media attribute, it is possible to supplementarily execute a process such as not mediating a book.

As described above, in the first embodiment, the user's preferences and the like are stored as profiles, and when a new user's profile is input, the profiles are collated, and the relevancy is determined using thesaurus knowledge. It can select and mediate high information.

(Embodiment 2) In Embodiment 1, when a new user's profile is input, only highly relevant information is selected. On the other hand, in Embodiment 2, the relationship is intentionally opposite. Only the information that is is selected.
This situation occurs when information that has a concept that is far from your usual taste, but information about another user who has similar characteristics to your own, is dared to be introduced. Used. This is effective, for example, when you are tired of a thrilling movie and want to watch a movie you want to relieve. In addition, there is research and demonstration data on the effectiveness of such an opposite concept and a heterogeneous concept. For example, a paper in the Journal of the Japanese Society for Artificial Intelligence Vol.11 No.6, “Examination of an information extraction method that has both relevance and heterogeneity for the purpose of supporting divergent thinking” is supported. The information intermediary device of the second embodiment has the same configuration as that of FIG. 1 except that only the function of the information intermediary unit 105 is different. The method of realizing the second embodiment differs from the first embodiment only in one place. That is,
Step 805 in FIG. 8 is changed to step 1005 in FIG. That is, it is determined whether or not the variables X and Y have ancestor concepts that conflict with each other. Here, the ancestor concept refers to a concept node that exists above a certain concept node. For example, the genre attribute of the rainman with ID = 17 is human touch, and S in the collation genre table.
F, Horror, and Hardboiled have their ancestor concepts at home and thrill suspense, respectively. Therefore, in the second embodiment, the rainman with ID = 17 is mediated by the user A. In FIG. 9, such a concept having the conflicting ancestors is referred to as “distant matching genre”.

On the other hand, if there is an ancestor concept that conflicts with each other, the process proceeds to step 806, and the variable XX is recorded in the mediation information output unit 107 as mediation information to the user A. If there is no ancestor concept that conflicts with each other, the process returns to step 803.

The second embodiment does not merely obtain the negated set of the first embodiment. It is highly surprising because the ancestor concept dares to mediate conflicting ones. By the way, the flaming dragon with ID = 18 has the genre attribute of kung fu,
It is neither a near match genre nor a far match genre. That is, no mediation is performed in this embodiment.

(Embodiment 3) The information intermediary apparatus of Embodiment 3 can handle not only a profile composed of information such as user's preference but also a content profile representing content information of a content such as a certain movie work. .

As shown in FIG. 11, the present information intermediary device
1 has the same configuration as that of FIG. 1 except that a content attribute information input unit 1108, a content attribute information group management unit 1109, a content attribute information classification information input unit 1110, and a content attribute information classification unit 1111 are added to the apparatus of FIG. ing.

The structure of the content attribute information will be described with reference to FIG. What is content attribute information?
This is information including an ID and an attribute of a certain content. For example, in the first of FIG. 12, the ID is 1201,
There is content attribute information in which the name attribute is a girl from Palestine, the genre attribute is hard boiled, and the media attribute is a book. Second, the ID is 1202
, But this is not related to the content attribute information with the ID 1201. That is, the order here has no particular meaning. For simplicity, it is assumed that the content attribute information in FIG. 12 has an attribute corresponding to one item of the profile (user profile) described in FIG. Of course, other attributes may be provided (for example, price for a book).
Others are the same.

Next, the operation of the present apparatus will be described with reference to FIGS. The plurality of pieces of content attribute information are input to the content attribute information group management unit 110 via the content attribute information input unit 1108 via a network or manually.
9 is stored. Information for classifying the content attribute information by a certain attribute is input from the content attribute information classification information input unit 1110 in FIG. 11 via a network or manually. For example, when media attribute = book is input, only the content attribute information having the media attribute of book is accumulated by the content attribute information classifying unit 1111 to form the profile shown in FIG. Of course, more book information will actually exist, so it is natural that various attribute information such as the year of issue, price, and author can be used in combination.

By treating this profile in the same way as the profile input from the user profile input unit 1101 and managing it in a unified manner by the profile group management unit 1102, the same operation and effect as in the first embodiment can be obtained.

The difference from the first embodiment is that in the first embodiment, if the user does not enter his / her preference information, the content attribute information does not exist and cannot be mediated. A new effect is that information from an input source other than the user, for example, information from a content distribution company can be input. In addition, the content attribute information can be mediated by the same configuration as the profile input from the user.

[0057]

As is apparent from the above description, the information mediating apparatus of the present invention accumulates user preference information and the like as a user profile, and when a new user profile is input, the information intermediary device between the user profiles is stored. By using the thesaurus knowledge, it is possible to select and mediate only those with high relevance from the collation results.

It is also possible to mediate information having a conceptually opposite relationship.

Further, when content attribute information is input, it is possible to classify the content attribute information according to a certain criterion and use it as a profile for mediation.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an information mediating device according to a first embodiment of the present invention.

FIG. 2 is a conceptual diagram showing an example of a profile.

FIG. 3 is a processing flowchart of the information mediating apparatus according to the first embodiment of the present invention.

FIG. 4 is a conceptual diagram illustrating a collation operation of the information mediating device according to the first embodiment of the present invention.

FIG. 5 is a conceptual diagram showing an example of a thesaurus of the information mediation apparatus according to the first embodiment of the present invention.

FIG. 6 is a processing flowchart for extracting a matching genre of the information mediating device according to the first embodiment of the present invention;

FIG. 7 is a conceptual diagram showing an example of a collation genre table of the information intermediation device according to the first embodiment of the present invention.

FIG. 8 is a flowchart of an information mediation process of the information mediation device according to the first embodiment of the present invention.

FIG. 9 is a conceptual diagram illustrating collation and mediation of the information mediating device according to the first embodiment of the present invention.

FIG. 10 is a flowchart showing a main part of an information mediation processing flow in a second embodiment of the present invention.

FIG. 11 is a block diagram showing a configuration of an information intermediary device according to a third embodiment of the present invention.

FIG. 12 is a conceptual diagram showing an example of content attribute information of an information mediating device according to a third embodiment of the present invention.

FIG. 13 is a conceptual diagram showing an example of a profile created from content attribute information of an information mediating device according to a third embodiment of the present invention.

[Explanation of symbols]

 101 Profile input unit 102, 1102 Profile group management unit 103, 1103 Profile collation unit 104, 1104 Collation result group holding unit 105, 1105 Information mediation unit 106, 1106 Thesaurus management unit 107, 1107 Mediation information output unit 1101 User profile input unit 1108 Content attribute information input unit 1109 Content attribute information group management unit 1110 Content attribute information classification information input unit

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Yoshihiro Noguchi 1006 Kazuma Kadoma, Kadoma City, Osaka Prefecture Inside Matsushita Electric Industrial Co., Ltd. 72) Inventor Takeshi Sakata 1006 Kadoma Kadoma, Kadoma City, Osaka Prefecture F-term in Matsushita Electric Industrial Co., Ltd. 5B049 AA01 5B075 NK46 PP02 PP03 PQ05 PR08 QM08 QP03

Claims (8)

[Claims] 1. A profile input unit for receiving a profile, a profile group management unit for receiving and managing profiles received from the profile input unit, and a profile group management unit for holding and managing profiles, and a profile group stored in the profile group management unit sequentially. A profile matching unit that extracts and performs matching on a plurality of profiles, a matching result group holding unit that receives information of a set of profiles matched by the profile matching unit and holds the result, and a matching stored in the matching result group holding unit An information intermediation unit that selects only highly relevant information from the collated profiles while referring to the thesaurus knowledge with the result as input, and an intermediation information output that transmits the intermediation information output from the information intermediation unit to the corresponding user An information intermediary device comprising: 2. The information intermediation device according to claim 1, wherein the information intermediation unit selects only information having the opposite relationship. 3. A content attribute information input unit for receiving content attribute information, a content attribute information group management unit for inputting and receiving and managing the content attribute information received by the content attribute information input unit, A content attribute information classification information input unit for inputting attribute information classification information, and a content attribute information classification for classifying content attribute information by inputting content attribute information classification information and information managed by a content attribute information group management unit The information mediating device according to claim 1, further comprising a unit. 4. A process of setting an input profile to a first variable, a step of checking whether there is an unchecked profile in the profile group management unit, and a step of checking if an unchecked profile exists. Taking out one and setting it to a second variable; and collating the first and second variables. When there is an item having the same content, the collation item is recorded and the first and second variables are recorded. Creating and storing a copy of the profile, returning the second profile to the profile group management unit when there is no collation item, and referring to the thesaurus knowledge when a collation profile pair exists. Selecting only highly relevant information. 5. The step of selecting only highly relevant information from the collated profiles includes the step of examining whether an unexamined item exists in the collated items, and the step of selecting whether the genre attribute of the item is A step of checking whether the genre attribute exists in the collation genre table; a step of adding the genre attribute to the collation genre table as a new collation genre if the genre attribute does not exist in the collation genre table; And performing a process of selecting mediation information when there is no more information. 6. The information mediation method according to claim 4, further comprising the step of: 6. The step of performing a process of selecting mediation information,
Checking whether an unchecked item exists in the items that have not been matched in the second profile; if so, the corresponding item as a third variable, and the genre attribute of the corresponding item as a fourth variable Setting; checking whether an untested item exists in the collation genre table; setting an genre attribute of the corresponding item as a fifth variable if an unchecked item exists; Check whether the 4th and 5th variables are the same concept. If they are not the same concept, check whether the 4th and 5th variables are sibling concepts or descendant concepts on the thesaurus. A step of confirming that the concept is not a concept, and if the fourth and fifth variables are the same concept, or are sibling concepts having a similar relationship, or a descendant concept. If you, the information intermediation method according to claim 5, characterized in that it comprises the step of selecting a third variable as an intermediary information. 7. The step of performing a process of selecting mediation information includes:
Checking whether an unchecked item exists in the items that have not been matched in the second profile; if so, the corresponding item as a third variable, and the genre attribute of the corresponding item as a fourth variable Setting; checking whether an untested item exists in the collation genre table; setting an genre attribute of the corresponding item as a fifth variable if an unchecked item exists; Determining whether the fourth and fifth variables have ancestral concepts that conflict with each other, and selecting the third variable as mediation information when the variables have ancestor concepts that conflict with each other. The information mediation method according to claim 5, wherein 8. A method comprising the steps of: inputting a plurality of pieces of content attribute information; and categorizing the plurality of pieces of content attribute information by a predetermined attribute, and processing the classified content attribute information as an input profile. The information mediation method according to claim 4, wherein: