JPH10232805A - Distributed retrieval system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress load on a retrieval device and a network and to reflect latest information on a retrieved result. SOLUTION: In the retrieval device, an input part 6 receives a retrieval request and a retrieval part 9 retrieves information pertinent to the retrieval request from an information supply device managed by the retrieval device. Then, an output part 11 informs a request origin of the retrieved result and further, the output part 11 propagates the retrieval request to the other retrieval device set as a propagation destination in a linking destination list 10. And a result, the retrieval request is propagated from one retrieval device to the other retrieval device and the request origin is informed of the retrieved results from the plural retrieval devices. Thus, since the retrieval request is delivered to the plural retrieval devices by the form of performing propagation, the load on the network is suppressed. Also, since retrieval is distributed by the plural retrieval devices, the load of the retrieval device is suppressed and the latest information is quickly retrieved.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a distributed retrieval system for retrieving information from a plurality of information supply devices connected via a network by using a plurality of retrieval devices. The present invention relates to a distributed search system for searching for information from existing information supply devices.

[0002]

2. Description of the Related Art Conventionally, by connecting a plurality of information supply devices holding information via a network,
A system for sharing information is known. In such a system, a search is performed to obtain information desired by a user from a large amount of information existing in the system.

A conventional example of such a search will be described below. For example, as a technology related to a search device for searching information (hereinafter, referred to as a first technology), a Lycos system of Carnegie Mellon University and a TITAN system of NTT Information and Communication Laboratories are known. This first technology is one of the information services on the wide area network, W
This is a technology for retrieving information desired by a user from the World Wide Web (WW). By reading all information resources that can be reached by following a reference relationship in advance and making the contents of the information resources into a database, Searches for the location of information that meets the user's request using the database in a short time.

As a technique for searching for information using a search device distributed in a network (hereinafter, referred to as a second technique), Japanese Patent Application Laid-Open No. Hei 6-110925, "Computers Connected by Network and How to Use Them" There is an invention described in ". This second technique broadcasts search keys to an unspecified number of search devices connected to a wide area network, receives search results from a plurality of search devices, and sorts, organizes, displays, and selects. By relaying to another network by broadcasting, it is possible to search for information resources existing on a plurality of networks.

Further, as a technique relating to a method or an apparatus for cooperating with a computer at a remote place (hereinafter referred to as a third technique), Japanese Patent Application Laid-Open No. 4-157944, entitled "Method and Apparatus for Controlling and Controlling Communication Network" 1), the invention described in Japanese Patent Application Laid-Open No. 1-194054, "Program Loading Method for Distributed Processing System" (hereinafter referred to as Document 2), and the invention described in Japanese Patent Application Laid-Open No. 7-6139, "Remote Procedure". Patent Document 3 (hereinafter referred to as Reference 3), Japanese Patent Application Laid-Open No. 7-18217.
There is an invention described in Japanese Patent Publication No. 4 “Remote programming method” (hereinafter referred to as Document 4).

In the invention described in Document 1, a program (agent) having a function of solving a partial problem for each of a plurality of computers connected to a network
Is placed. When a certain problem is given, the problem is divided into a plurality of sub-problems and assigned to an appropriate agent on the network. In the invention described in Reference 2, each agent obtains a solution cooperatively via a network. In an environment in which a plurality of computers are connected to a network, a failure in a program running on a certain computer may occur. Request another computer to provide a program required when a problem occurs. Then, the computer that has received the request to provide the program provides the program to the request source according to the content of the request, and the request source selects and uses the program considered most appropriate for the request.

According to the invention described in Document 3, in an environment where a plurality of computers are connected to a network, when a remote procedure call (RPC) is propagated to each computer and executed in a nested form, during execution of the RPC, When the error occurs, the computer resources of the caller are immediately released. In the invention described in Document 4, processing is performed while mutually exchanging programs between processors on a network, and the programs can be interpreted by all processors on the network.
The program is transferred to another processor by an instruction called o, and the program is executed in a chain.

[0008]

According to the first technique described above, it is necessary for the search device to acquire information in advance, summarize the contents of the information and hold it in a database in the device itself. There was a problem that as it became larger, it had to have a large database. Further, in order to search for information desired by the user from the database, when the location of the searched information has already become invalid, or when new information appears, etc.,
There have been problems in that the user is shown information that has already been unusable or cannot know the existence of the latest information. This problem can be solved by newly acquiring information and reflecting it in the database. However, as the amount of information becomes large, it takes a long time to reflect the information and the effect cannot be expected. . In addition, since a single search device provides services to a large number of users, there has been a problem that the load on the search device becomes extremely large.

According to the second technology, search devices are distributed and arranged on a wide area network to reduce the load on each search device. However, a search request to each search device is broadcast. , There is a problem that the load on the network becomes very large. The third technology described above is a component technology for solving a problem by cooperative operation of distributed processing entities (agents) on a network, and a distributed search device is configured using these component technologies. However, when such a configuration is adopted, it is necessary to negotiate cooperative operations between distributed processing entities and to deliver programs, and to store information other than information necessary for searching. There is a problem that a load is imposed on the network due to the necessity of communication.

The present invention has been made in view of the above-mentioned conventional circumstances, and provides a distributed search system capable of suppressing the load on a search device and a network and reflecting the latest information in search results. It is intended to be.

[0011]

To achieve the above object, a distributed search system according to the present invention comprises a plurality of information supply devices for holding information and a plurality of search devices for retrieving information from the information supply devices. And a search device connected via a network, the search device comprising: an input unit that receives a search request; and a search unit that searches for information corresponding to the search request from an information supply device managed by the search device. And a notifying means for notifying a search result to a request source, and a propagating means for propagating the search request to another search device set in advance as a propagation destination.

In the above-described distributed search system, in the search device, the input means receives the search request, and the search means searches the information supply device managed by the search device for information corresponding to the search request. Then, the notifying unit notifies the request source of the search result, and the propagation unit propagates the search request to another search device set in advance as a propagation destination. As a result, a search request is propagated from one search device to another search device, and the request source is notified of search results from a plurality of search devices.

As described above, since a search request can be transmitted to a plurality of search devices by propagating the search request, compared with the case of performing broadcast communication over the entire network or the case of transferring a program between devices, Thus, the burden on the network can be suppressed. In addition, by distributing and managing a plurality of information supply devices to a plurality of search devices, the burden on each search device can be suppressed, so that information can be searched quickly, and the latest information can be retrieved. Can be easily reflected.
The information supply device managed by each search device is preferably an information supply device that is close to the search device on a network basis.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS As a distributed search system according to a first embodiment of the present invention, an HTTP (hypertext transport
A distributed search system in the WWW (Web) for exchanging information according to a protocol (rotocol) will be described with reference to FIG.
FIG. 1 shows a part of a distributed search system. The distributed search system includes a plurality of Web servers 1 for storing information (pages), a plurality of news servers 2 for storing a plurality of information (news articles), and a Web browser 3 for receiving search conditions from a user. A plurality of search devices 4 for searching for information; the Web server 1, the news server 2, and the Web browser 3
It is connected via the Internet 5. The search device 4 is connected to the Web server 1. It should be noted that ranges A and B surrounded by broken lines in the drawing indicate ranges in which one search device 4 searches for information.

The Web browser 3 receives a search instruction for desired information from a user, converts the received instruction into a message representing a search request, and transmits the message to the search device 4. When a message containing the search result is received, the search result is extracted from the message and displayed.

The above-described search instruction is received by, for example, the interface shown in FIG. 2. The search instruction includes a search expression representing desired information, a search type, a search target, a return address of a search result, and the like. is there. As a search expression, for example, a URL (Universal Resource Locato
r), keyword, combination of multiple keywords, combination of URL and keyword, multiple U
An RL regular expression connected by a set operator (set product, set sum, set difference, etc.) is received. The keyword does not mean one word, for example,
It also means multiple words. As the types of search, one of a normal search in which the search is performed only once from the search target, a latest information search in which the latest information is sequentially searched from the search target, and a deletion in which the latest information search is stopped is accepted. As a search target, any of Web search (page search), news search (news article search), and both (page and news search) is accepted.

Here, the above-mentioned search instruction is converted by the Web browser 3 into a message having the following data structure. In the present embodiment, a message in a common format is used not only in a case where a search instruction is converted but also in another process of another device. As shown in FIG. 3, this message includes a “message ID”, a “search formula”,
"Search target", "Search type", "Search history"
"Search result".

The "message ID" is an identifier that uniquely represents each message. When the message is generated, for example, the time when the message is generated, the name of the host that generates the message, the process ID number when the message is generated, etc. Is set based on It should be noted that an identifier indicating the message to be deleted is set in the deletion message for stopping the latest information search. The “search formula” is a search formula for information to be searched. “Search target” is a code indicating the search target, and “Web” in the case of a Web search.
However, “news” is set in the case of the news search, and “Web, news” is set in the case of the Web search and the news search. The “search type” is a code indicating the search type, and “common” is set for a normal search, “new” for a latest information search, and “delete” for a deletion. The “search history” is a history in which the IDs of search devices that have performed a search based on the message are sequentially displayed. At the top of the “search history”, the mail address of the request source that generated the message is set. The “search result” is a search result for the message. For example, in the case of a Web search, the URL of the information, the title of the information, and the last information update date and time are set. The URL of the information and the ID of the news are set.

Next, a search device 4 which is a central part of the present invention.
Will be described with reference to FIG. The search device 4 includes an input unit 6, an interpreter unit 7, a latest information search database 8, a search unit 9, a link destination list 10, and an output unit 11.
And The input unit 6 receives messages and accumulates the messages in the order in which the messages were received.

The interpreter unit 7 has a function of reading a message stored in the input unit 6, a function of detecting a "search type" of the read message, and a "search type" in the read message that is "latest information search". And a function to copy the message based on the fact that it is stored in the latest information search database, and delete the corresponding message from the latest information search database based on the fact that the "search type" in the read message is "delete" And a function of transferring the message to the output unit 11 and a function of transferring the message to the search unit 9.

The latest information search database 8 holds a message in the data format shown in FIG. 3 including a search request for the latest information search. The search unit 9 has a search function of searching for information based on a message from the Web server 1 and the news server 2 within a range managed by the search device. When a message is received from the interpreter unit 7, a search function is performed based on the message. In the function of causing the search function to perform a search, the function of reading a message from the latest information search database 8 at predetermined time intervals, and the function of causing the search function to search based on the message, and the search for the message read from the latest information search database 8, A function of searching for information different from the previous search result, a function of reflecting a search result in a message of the latest information search database 8 when a search is performed on a message read from the latest information search database 8, and a message of the search result And a function of transferring to the output unit 11 It has.

The search function includes a function of extracting a "search formula" from a message, a function of extracting a "search target" from a message, and a Web server 1 belonging to a management range.
A web search unit 9a for searching for a news server 2, a news search unit 9b for searching for a news server 2 belonging to a range to be managed,
A function of causing the Web search unit 9a and the news search unit 9b to search according to the "search target".

The link destination list 10 is the Internet 5
The host names of other search devices 4 connected via the server are stored. For example, the search device 4 belonging to the range A in FIG. 1 stores the host name of the search device 4 belonging to the range B.
Note that the number of search devices to be propagated is not limited to one.

The output unit 11 outputs the "search history" of the message.
, A function of propagating a message including the search request received from the interpreter unit 7 to another search device 4 based on the link destination list 10, and a search result received from the search unit 9. A function of notifying the message to the search request source.

The input means referred to in the claims is mainly constituted by the input unit 6, the search means and the re-search means are mainly constituted by the search unit 9, and the holding means and the search result holding means are , Is mainly configured by the latest information search database 8, and the notification unit, the propagation unit, the search restriction unit, and the re-notification unit are mainly configured by the output unit 11.

Next, the operation of the above-mentioned distributed search system will be described focusing on the operation of the search device 4. Note that communication via the Internet 5 is generally known, and a description thereof will be omitted. First, when the input unit 6 receives a message via the Internet 5, the interpreter unit 7 performs the processing shown in FIG. That is, the interpreter unit 7 sequentially retrieves the messages stored in the input unit 6 (step S1), and detects whether or not the “retrieval type” of the retrieved message is “latest information retrieval” (step S2). In the case of "latest information search", a copy of the message is created and the latest information search database 8 is created.
(Step S3) On the other hand, if it is not "latest information search", nothing is performed. The message stored in the latest information search database 8 is searched for the latest information by the search unit 9 as described later.

Next, it is detected whether or not the "search type" of the retrieved message is "delete" (step S).
4) In the case of "deletion", it is further detected whether or not a message having the same message ID as the extracted message is stored in the latest information search database 8 (step S5). If there is, the corresponding message is deleted from the latest information search database 8, and the message is excluded from the target of the latest information search processing described later (step S6).

Then, the extracted message is transferred to the output unit 11 (step S7), and the message whose "search type" is other than "delete" is further transferred to the search unit 9 (step S8). The message transferred to the output unit 11 is propagated to another search device 4 by the processing of the output unit 11 as described later, and the message transferred to the search unit 9 is processed by the search unit 9 as described later. Search processing is performed.

Upon receiving the message from the interpreter unit 7 as described above, the search unit 9 performs a search process shown in FIG. That is, the search unit 9 extracts a “search formula” from the message received from the interpreter unit 7 (step S9), extracts a URL or a set of keywords from the extracted “search formula” (step S10), and The search target is extracted (step S11). Then, it is detected whether or not “Web” is included in the “search target” in the message (step S12), and “Web” is detected.
Is included, the Web server 1 is searched based on the search formula to obtain the URL of the corresponding information, the title, and the date and time of the last information update (step S13). It is detected whether “news” is included in the “target” (step S14).

If "Web" is not included in the search target in the message or "news" is included, the news server 2 is searched based on the "search formula", and URL of information and news I
D is acquired (step S15).

Then, "ne" is set as a search target in the message.
If “ws” is not included, or if information is searched from the news server 2, a search result is generated by performing a set operation on the URL obtained by the search based on the “search formula”. (Step S16). Next, a search result is set in the message (step S1).
7), the message is transferred to the output unit 11 (step S18). This message is transmitted by the output unit 11 to the request source that created the message, as described later.

Further, the retrieval section 9 reads the message stored in the latest information retrieval database 8 by the interpreter section 7 as described above, and performs the latest information retrieval processing shown in FIG. The search unit 9 detects whether or not a certain time has elapsed by using a timer (not shown) (step S1).
9) When the predetermined time has elapsed, the following processing (step S
20 to S30). Note that, even if a certain time has elapsed, if the above-described normal search processing is being performed, the processing is executed after the end of the normal search. First, a message is read from the latest information search database 8, and a "search result" is extracted from the message (step S20). Next, a “search expression” is extracted from the message (step S21), and UR is extracted from the extracted “search expression”.
L or a set of keywords is extracted (step S2).
2) Further, “search target” is extracted from the message (step S23). Then, it is detected whether or not "Web" is included in the "search target" in the message (step S24). If "Web" is included, the Web server 1 is searched based on the search formula. Then, the update date of the retrieved information is compared with the last information update date and time of the same information in the extracted “search result”, and the URL, the title, and the last information update date and time of the information having different dates and times are obtained ( Step S25). As a result, information having the same content as before can be excluded from the search result, and only the latest information can be obtained. Next, “n” is set as “search target” in the message.
ews "is detected (step S
26).

If "Web" is not included in the search target in the message or "news" is included, the news server 2 is searched based on the "search formula", and URL of information and ID of news
Is acquired (step S27).

Then, "ne" is set as a search target in the message.
If “ws” is not included, or if information is searched from the news server 2, a search result is generated by performing a set operation on the URL obtained by the search based on the “search formula”. (Step S28).

Next, the search result is set in the read message (step S29), the message is transferred to the output unit 11 (step S30), and the search result is reflected on the message in the latest information search database 8. The message transferred to the output unit 11 is transmitted by the output unit 11 to the request source that created the message, as described later.

As described above, upon receiving a message from the interpreter unit 7 or the search unit 9, the output unit 11 performs the processing shown in FIG. That is, the output unit 11 analyzes the number of searches performed by another search device from the “search history” of the received message (step S31), and determines whether the number is less than a predetermined number (for example, 10). Is detected (step S32).

As a result, if the number is equal to or more than the predetermined number, it means that an appropriate search is being performed, so the message is discarded, and the process is terminated. This prevents a useless message from being transmitted over the network, and also prevents the use of another search device in which a process based on the message is continuously performed. On the other hand, if the number is less than the predetermined number, it is detected whether the received message is a message received from the interpreter unit 7 (step S34). If the received message is a message received from the interpreter unit 7, the message is Since this means that the message is transmitted to another search device, the ID of the search device is included in the “search history” in the message.
Is added (step S35), the host name of the other search device of the propagation destination is acquired from the link destination list 10 (step S36), and the message is transmitted to the relevant search device 4 (step S37). As a result, the processing for the message is performed in the other search devices 4 as well.

On the other hand, if the message is not a message received from the interpreter unit 7, that is, a message received from the search unit 9, it is a message including a search result. The request source mail address set at the top is obtained (step S38), the "search result" of the message is encrypted (step S39), and the message is transmitted to the address by e-mail (step S40). As a result, the search result is transmitted to the request source via the Internet 5.

Therefore, the burden on the network can be reduced and the load on each search device can be reduced as compared with the case of performing the broadcast communication and the transfer of the program between the devices. The search can be performed quickly, and the latest information can be easily reflected in the search result. Also, desired latest information can be sequentially provided to the request source.

Next, a distributed search system according to a second embodiment of the present invention will be described with reference to the drawings. In the following description, the same parts as those in the first embodiment are denoted by the same reference numerals, and the duplicate description will be omitted. This distributed search system is configured to further add related peripheral information to the search result to the distributed search system of the first embodiment described above. Are different from each other.

This search device 12 has a function to be described later added to the search device 4 of the first embodiment, a peripheral information database 13 as a peripheral information holding means, a peripheral information management unit 14, and an output unit 11. An output unit 15. The peripheral information database 13 stores a keyword and another keyword (peripheral information keyword) related to the keyword.
And the frequency at which the peripheral information keyword appears in the search formula in the message, and the message ID of the message generated when searching for the peripheral information keyword, as described later, are held in association with each other.

The peripheral information management unit 14 has a management function for managing the peripheral information database 14. The details of the management function will be described later in the description of operation.

The output unit 15 acquires, from the peripheral information database 14, a peripheral information keyword related to the keyword of the search expression included in the message received from the search unit 9, in addition to the function of the output unit 11. A function to add to the search results, a function to create a message when the latest information search is performed by the peripheral information keyword, a function to create a deletion message to end the latest information search for the peripheral information keyword by another search device, And a function of transmitting the created message to another search device based on the link destination list 10.

Here, the information adding means and the second propagation means referred to in the claims are mainly constituted by the output unit 15, and the registration means, the reception means and the second registration means are mainly constituted by the peripheral information management unit 13. The search request creation means is mainly configured by the peripheral information management unit 13 and the output unit 15.

Next, the operation of the distributed search system according to the second embodiment will be described with reference to the drawings. FIG. 10 shows a peripheral information adding process for adding related peripheral information to a search result.
This will be described with reference to FIG. This peripheral information adding process is performed by the output unit 15, and is a process in which step S41 is added to the process of the output unit 11 shown in FIG. That is, the output unit 15 extracts a keyword from the “search formula” of the message received from the search unit 9, acquires a peripheral information keyword corresponding to the keyword from the peripheral information database 14, and outputs the keyword to the “search result” of the message. Add.

Accordingly, the search result is transmitted to the request source, and, for example, when a search relating to “computer” is performed, the peripheral information keywords such as “hard disk” and “network” are also transmitted, and the search is performed. It is possible to provide the request source with a new and highly relevant viewpoint that was not intended at the time.

Next, the peripheral information management processing by the peripheral information management unit 13 will be described with reference to FIG. The peripheral information management unit 13 monitors the “search formula” of the message received by the output unit 15, detects a search keyword that appears frequently (step S42), and appears simultaneously with the search keyword from the “search result” of the message. A peripheral information keyword that is frequently performed is detected (step S43), and it is detected whether a set of the search keyword and the peripheral information keyword is registered in the peripheral information database 14 (step S44).

As a result, if not registered, a set of the search keyword and the peripheral information keyword is registered in the peripheral information database 14 (step S45). Then, a search formula is generated based on the peripheral information keywords. Note that this search formula may be a single keyword, a set sum of a plurality of keywords, or a formula using various other rules of thumb. Then, the generated search formula is output to the output unit 1.
5 and receives the message ID of the message created by the output unit 15 as described later, and stores the message ID in the peripheral information database 14 in association with the peripheral information keyword (step S46). If it is registered, nothing is done.

Next, when an e-mail including a search result for the message is received, another peripheral information keyword is extracted from the search result and stored in the peripheral information database 14 in association with the search keyword (step). S47, S48). Then, a peripheral information keyword whose detection frequency is equal to or lower than a predetermined frequency is detected from the peripheral information database 14, a message ID corresponding to the peripheral information keyword is obtained, and the message ID is passed to the output unit 15 to output the latest information. An instruction to delete the search message is issued, and the peripheral information keyword is deleted from the peripheral information database 14 (steps S49 and S50).

When the search formula is received from the peripheral information management unit 13 as described above, the output unit 15 uses the search formula to generate the search formula shown in FIG.
2 is performed. That is, the output unit 15 receives the search formula from the peripheral information management unit 13 (Step S5).
1), "Web, news" is set in the "search target",
Set “new” in “Search type” to indicate the latest information search,
A message in which the ID of its own search device is set in the "search history" is generated (step S52). The message ID of the generated message is notified to the peripheral information management unit 13. Then, the host name of another search device of the propagation destination is acquired from the link destination list 10 (step S53), and the message is propagated to the corresponding search device 4 (step S53).
54). As a result, the latest information search is performed in the other search device 4, and the search result is returned to its own search device.

As described above, the instruction to delete the latest information search message from the peripheral information management unit 13 and the message I
When D has been acquired, the output unit 15 performs the latest information search message deletion process shown in FIG. That is,
When receiving the latest information search message deletion instruction and the message ID (step S55), the output unit 15 generates a latest information search deletion message corresponding to the received message ID (step S56), and transmits the other search device to the propagation destination. From the linked list 10
The message is propagated to the corresponding search device 4 (step S57). As a result, the message of the latest information search corresponding to the message ID is deleted in another search device 4, and the latest information search ends.

Therefore, not only the search result but also the peripheral information related to the search can be provided to the request source.
Further, the search device itself can collect the peripheral information from another search device, and the peripheral information provided to the request source can be enriched.

In the above-described embodiment, since the search device directly searches for information at the time of receiving the search request, the search device can perform the search in which the latest information is reflected, and can perform the search with the storage device and the like. Although the control of the resources has been performed more effectively, the present invention is not limited to this. For example, the gist of the information may be managed in advance as a database, and the database may be searched. Further, in the above-described embodiment, each search device returns the search result directly to the request source. However, the search result is propagated in the same manner as a message, and the search results obtained by a plurality of search devices are collected. It may be returned to the request source. Further, in the above-described embodiment, a message transmission destination that is managed in advance by the search device is used. For example, information on the transmission destination is set in the search request,
The message may be propagated based on the information of the propagation destination. The distributed search system according to the present invention can be used in various information searches, such as searching various information resources on the Internet, searching business-related information in an intranet, and the like.

Each of the above functional units is constituted by a processor executing a control program (prestored in a ROM, a RAM or the like). In the present invention, these functional units are constituted as independent hardware circuits. May be. Further, the present invention executes the processing according to the present invention by causing the control program to be grasped as a storage medium such as a floppy disk or a CD-ROM, and inputting the control program from the storage medium and causing the processor to execute the control program. It can also be done.

[0055]

As described above, in the distributed search system according to the present invention, since a search request is transmitted to a plurality of search devices in a mode of propagating the search request, a broadcast communication is performed for the entire network. And the burden on the network can be suppressed as compared with the case where a program is transferred between devices. In addition, by distributing and managing a plurality of information supply devices to a plurality of search devices, the burden on each search device can be suppressed, so that information can be searched quickly, and the latest information can be retrieved. Can be easily reflected. Furthermore,
In the distributed search system according to the present invention, since the search request is held and the search is sequentially performed, the request source can be quickly notified of the latest information.

[Brief description of the drawings]

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

FIG. 2 is an example of a user interface for inputting search conditions according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating a data format of a message according to the embodiment of the present invention.

FIG. 4 is a configuration diagram of a search device according to the first embodiment of the present invention.

FIG. 5 is a flowchart illustrating an operation of the interpreter unit according to the first embodiment of the present invention.

FIG. 6 is a flowchart illustrating an operation of a search process of a search unit according to the first embodiment of the present invention.

FIG. 7 is a flowchart illustrating an operation of a latest information search process of a search unit according to the first embodiment of the present invention.

FIG. 8 is a flowchart illustrating an operation of the output unit according to the first embodiment of the present invention.

FIG. 9 is a configuration diagram of a search device according to a second embodiment of the present invention.

FIG. 10 is a first flowchart illustrating the operation of an output unit according to a second embodiment of the present invention.

FIG. 11 is a flowchart illustrating an operation of a peripheral information management unit according to a second embodiment of the present invention.

FIG. 12 is a second flowchart illustrating the operation of the output unit according to the second embodiment of the present invention.

FIG. 13 is a third flowchart illustrating the operation of the output unit according to the second embodiment of the present invention.

[Explanation of symbols]

1. Web server, 2. News server, 3. Web browser, 4, 12
・ Search device, 5 ・ Internet, 6.
-Input part, 7-Interpreter part, 8-
Latest information search database, 9 ... search section,
10 link list, 11, 15, output section, 13 peripheral information management section, 14
A peripheral information database.

Claims (7)

[Claims] 1. A distributed search system in which a plurality of information supply devices for holding information and a plurality of search devices for searching for information from the information supply device are connected via a network, wherein the search device comprises: An input unit for receiving a request, a search unit for searching for information corresponding to the search request from an information supply device managed by the search device, a notification unit for notifying a search result to a request source, and a notification unit set in advance as a propagation destination. And a propagation means for propagating the search request to the search device of (1). 2. The distributed search system according to claim 1, wherein the search request includes a search history, and the search device further terminates execution of the search processing based on the search history. A distributed search system comprising search restriction means. 3. The distributed search system according to claim 1, wherein said search device further comprises: holding means for holding the received search request; and information based on the held search request. A distributed search system comprising: re-search means for re-searching applicable information from a supply device; and re-notification means for notifying a request source of a re-search result. 4. The distributed search system according to claim 3, wherein the search device further includes a search result holding unit that holds a search result obtained by a search unit, and the re-search unit includes an information supply device. A distributed search system for searching for updated information in information corresponding to a held search result. 5. The distributed search system according to claim 1, wherein the search device further stores a keyword and a keyword related to the keyword in association with each other. A distributed search system comprising: a holding unit; and an information adding unit that adds a keyword corresponding to a keyword included in a search request to at least one of a search result and a re-search result. 6. The distributed search system according to claim 5, wherein the search device further associates a keyword included in the search request with a keyword included in a search result with respect to the search request and associates the keyword with the keyword included in the search result. A distributed search system, characterized by comprising a registration means for registering in a database. 7. A distributed search system according to claim 6, wherein: a search request creating means for creating a peripheral information search request based on a keyword included in a search result for the search request; Second transmitting means for transmitting the peripheral information search request created to the search device of the above, receiving means for receiving a search result for the peripheral information search request, and a keyword included in the received search result and a keyword included in the search request. And a second registration means for registering the information in the peripheral information holding means in association with each other.