JP2008140283A - Distributed hierarchy-based information search method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve such a problem that in an existing search system, only the terminals which can be connected directly through a network can be searched. <P>SOLUTION: Nodes to be searched are hierarchized and a node list server having management information of the node of each layer is installed in each hierarchy. When searching, a search agent requests search to a primary node. The primary node starts to search its own node, and simultaneously, if a secondary node is bound, the primary node requests to search the secondary node. Thereafter, the n-th node starts to search its own node, and simultaneously, if the (n+1)-th node is bound, the n-th node requests to search the (n+1)-th node. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an information search method in a distributed hierarchical base type system configuration.

  The main object of the present invention is to search for information managed by each company across the board. The searched information is “service” information such as products handled in each company, and it is appropriate to manage the information at the data generation source in consideration of the management cost of the information.

  A conventional general information search system is a method for performing information aggregation and search indexing. With this method, it is relatively easy to consolidate resources within a single company, but when searching for each company across the board, a joint investment company or a third party is required, and the investment scale is large. It became a thing. Also, such a system took too long to build and as a result it was sometimes useless.

  In the present invention, the exchange of information is ruled, and an independent infrastructure system is constructed that does not depend on the implementation in the company (node). As a result, each company can build a system that provides timely information with small resources.

JP-A-6-301720 JP 2003-308321 A

  None of the above prior arts perform a search across a plurality of different companies divided in a network.

  The present invention has been made in view of the above points, and an object of the present invention is to provide an information search method in a distributed hierarchical base type system configuration across a plurality of different companies.

  The nodes that execute the search are hierarchized, and a node list server having management information of the nodes in the same hierarchy is set in each hierarchy. When performing a search, the search agent requests the primary node to perform the search. The primary node starts searching for its own node, and at the same time, if the secondary node is bound, requests the secondary node to search. Thereafter, the n-order node starts searching for its own node, and at the same time, if the n + 1-order node is bound, requests the n + 1-order node to search.

By constructing the above system, the following merits can be obtained.
Search can be performed without constructing a large-scale search system.

  The search agent, node list server, and node can be configured with inexpensive terminals.

  The nodes are hierarchically managed, and the increase / decrease of the nodes only needs to be grasped by the node list server of that hierarchy, and it does not affect other hierarchies. That is, the agent only needs to request a search to the primary node, and even if the number of nodes lower than the secondary is increased or decreased, the agent search method is not affected.

  According to the present invention, information stored in each node (terminal) can be searched in an integrated manner in a network-divided environment.

  Further, according to the present invention, it is possible to provide an information search method in a distributed hierarchical base type system configuration across a plurality of different companies.

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

  In this embodiment, a search agent, a primary node list server, and a primary node are connected through an Internet line, and a plurality of secondary nodes are linked to one primary node through an Internet line or a LAN (local area network) line. An example will be described.

  FIG. 1 is a schematic diagram of an embodiment of the present invention. In FIG. 1, 10 is a client terminal, 20 is a search agent, 30 is an Internet line, 40 is a primary node list server, and 41, 42 and 43 are primary nodes. 50 is a local area network, 60 is a secondary node list server, 61, 62, 63 and 64 are secondary nodes, 70 is an n-order node list server, and 71 is an n-order node.

  2 is a hardware block diagram of this implementation example, FIG. 3 is a table structure diagram for managing the nodes stored in the node list server, FIG. 4 is a chart diagram during search execution, and FIG. 5 is an implementation of the present invention in the travel industry. For example, FIG. 6 is a table structure diagram of ID management means stored in the primary node list server, and FIG. 7 is a table structure diagram of condition determination means possessed by the node.

  When performing a search, the user transmits authentication information to the search agent 20 through the client terminal 10 and performs a search by specifying a search condition. The search agent 20 inquires of the primary node list server 40 for primary node information and confirms user authentication information via the Internet line 30. The primary node list server 40 searches the node management table shown in FIG. 3 and the ID management means table shown in FIG. 6 and answers the search agent 20 with the confirmation information of the primary node information and the user authentication information. The search agent 20 uses the information to make a search request for the primary node list 40, 41, 42.

  In the primary nodes 41, 42, and 43, first, a search in the own node is executed. In the primary nodes 42 and 43 to which the secondary node is not tied, the search result in the own node is returned to the search agent 20 and the search process is terminated. The primary node 41 associated with the secondary node performs a search in its own node and simultaneously makes an inquiry for secondary node information to the secondary node list server 60 via the local area network 50. The secondary node list server 60 returns the secondary node information to the primary node 41. Using the information, the primary node 41 makes a search request to the secondary nodes 61, 62, 63, and 64.

  Note that the secondary node is not necessarily located on the local area network, and may be located on the Internet like the secondary node 61.

  In the secondary nodes 61, 62, 63, and 64, the search in the own node is first executed. In the secondary nodes 61, 62, and 63 to which the lower nodes are not bound, the search result in the own node is answered to the primary node 41, and the search process is terminated. Since the secondary node 64 is associated with lower nodes, the secondary node 64 performs a search in its own node, and simultaneously inquires about lower node information and makes a search request to the lower node.

  Thereafter, the above processing is repeated and a search request is performed up to the n-th node.

  Search results obtained at the secondary nodes 61, 62, 63, 64 are collected at the primary node 41. The primary node 41 transmits the search results to the search agent 20 at intervals of a predetermined time even if the search results of all lower nodes are not obtained.

  The search agent 20 transmits the current search result to the client terminal 10 with a predetermined time interval. With such a mechanism, the user can refer to the search result of each node from the client terminal 10.

  In addition, when the node search process is completed, the node transmits end information to the upper node. In the case of a node to which a lower node is associated, the end information is transmitted to the upper node when the search process of the own node is completed and the completion information is received from all the associated lower nodes. In the above process, the end information is collected in the upper node, and the entire search process ends when the search agent 20 receives the end information from all the primary nodes.

  If the user obtains necessary information before the entire search process is completed or if the user wants to end the search process, the search is stopped by operating the client terminal 10. In that case, the search agent 20 transmits a search stop request to the primary nodes 41, 42, 43. The primary nodes 41, 42, and 43 that have received the search stop request stop the search process for the own node. In addition, the primary node 41 to which the secondary node is linked transmits a search stop request to the secondary nodes 62, 63, and 64. Thereafter, the above processing is repeated, and a search stop request is performed up to the n-th node.

  Using the hardware block diagram of this implementation example of FIG. 2, the process from when the search agent makes a search request to obtaining the search result is shown. When the client receiving unit 101 of the search agent 20 receives the authentication information and the search information (eg, region: travel to the United States: 3/21 to 4/1, budget: 200,000 to 300,000) from the client terminal, the client receiving unit 101 Pass the search condition. The lower node management unit 103 includes destination information of the primary node list server, and makes an inquiry to the node management unit 121 and the ID management unit 122 of the primary node list server 40 via the Internet line 30. The node management unit 121 transmits primary node information to the lower node management unit 103 of the search agent 20. The ID management unit 122 checks the validity of the ID input by the user using the PASSWORD or the like, and transmits the user authentication information to the lower node management unit 103 of the search agent 20.

  The lower node management means 103 makes a search request to the search request receiving means 131 of the primary node 41 based on the obtained primary node information. The search request is generated in the form of a query based on the conditions specified by the user on the GUI of the client terminal and the user ID. The search request receiving unit 131 interprets the search request and passes the search condition to the search execution unit 132. The search execution means 132 starts searching the information DB 136 of the own node 41 and requests the lower node management means 133 to search for lower nodes. The lower node management unit 133 makes an inquiry to the node management unit 141 of the secondary node list server 140 via the local area network 50. The node management module 141 returns secondary node information to the lower node management means 133.

  The lower node management means 133 makes a search request to the search request receiving means 151 of the secondary node 64 based on the obtained secondary node information. The search request here is the same as that executed from the search agent subordinate node 103 to the search request receiving means 131 of the primary node 41. The search request receiving unit 151 passes search conditions to the search execution unit 152. The search execution unit 152 starts searching the information DB 156 of the own node 64 and at the same time requests the lower node management unit 153 to search for the lower node.

  The above series of processing is repeated to reach the lowest n-th order node list server 70 and n-th order node 71, and the search request processing ends.

  The search result collection unit 154 of the secondary node 64 collects the search results from its own node and lower nodes, and makes an inquiry to the condition determination unit 155. The condition determining means 155 returns the answer condition to the search result collecting means 154 based on the user authentication information, and the search result collecting means 154 answers to the search result collecting means 134 of the primary node 41 based on it.

  The search result collection unit 134 of the primary node 41 collects the search results from its own node and lower nodes, and makes an inquiry to the condition determination unit 135. The condition determining means 135 returns the answer condition to the search result collecting means 134 based on the user authentication information, and the search result collecting means 134 answers to the search result collecting means 104 of the search agent 20 based on it.

  The search result collection unit 104 of the search agent 20 receives the search result transmitted from the search result collection unit 134 of the primary node 41 with a predetermined time interval, and passes it to the client transmission unit 102. The client transmission unit 102 transmits the search result to the client terminal. The search result collection unit 134 of the primary node 41 transmits the end information to the search result collection unit 104 of the search agent 20 when the search of the own node and the lower node is completed. When the search result collection unit 104 passes the end information to the client transmission unit 102, the search process ends.

  FIG. 3 is a table structure diagram of node information stored in the node list server. Information on the n-th order node is stored in the n-th order node list server. The stored information includes a node number 200 of each node and URL information 201 indicating each node. In the URL information, the position information of the node on the network can be uniquely specified. In some cases, a DNS server is used for location resolution, but it uses an existing general mechanism.

  FIG. 4 is a flowchart from the start of the search by the search agent to the end of the search. When the search agent 20 in FIG. 1 receives the search condition from the client terminal 10, it makes an inquiry about the primary node information to the primary node list server 40. The primary node list server 40 provides primary node information to the search agent 20 (step 300).

  The search agent 20 issues a search request to the primary nodes 41, 42, and 43 based on the primary node information (step 301).

  The primary nodes 42 and 43 search their own nodes and return the results to the search agent 20. Since the secondary node is bound to the primary node 41, the secondary node list server 60 is inquired of the secondary node information at the same time as the search for the local node is started. The secondary node list server 60 provides primary node information to the primary node 41 (step 302).

  The primary node 41 makes a search request to the secondary nodes 61, 62, 63, and 64 based on the secondary node information (step 303).

  The secondary nodes 61, 62, and 63 perform their own node search and return the result to the primary node 41. Since the secondary node 64 is associated with lower nodes, a search for lower nodes is also performed.

  The above-described series of processing is repeated to reach the lowest n-th order node list 70 and n-th order node 71, and the search results are finished. Search results are collected in the secondary node 64 (steps 304, 305, 306).

  The secondary node 64 transmits the search result of its own node and lower nodes to the primary node 41 (step 307).

  The primary node 41 transmits the search result of its own node and lower nodes to the search agent 20 (step 308).

  The search result is passed from the search agent 20 to the client terminal 10 and presented to the user.

  FIG. 5 shows an implementation example in the travel industry of the present invention. The travel industry association has a primary node list server, A travel agency and B travel agency have client terminals respectively, C travel planning companies and D travel planning companies have primary nodes and their primary nodes. There are subordinate nodes attached.

  Here, it is assumed that a search request is made to the search agent 410 from the client terminal 400 of the A travel agency. The search agent 410 makes an inquiry to a temporary node list server installed in the travel industry association, and obtains location information and client information of the primary nodes 430 and 440. Several means for discriminating the client information are conceivable. Here, as shown in FIG. 6, it is assumed that the discriminating is based on the ID and PASSWORD.

  The search agent 410 makes a search request to the primary nodes 430 and 440. The primary node 430 starts a search for its own node and simultaneously makes an inquiry to the secondary node list 431 to obtain position information of the secondary nodes 432 and 433. Based on this, the primary node 430 makes a search request to the secondary nodes 432 and 433. Each of the secondary nodes 432 and 433 searches for its own node and returns the result to the primary node 430. The primary node 430 sends the search result to the search agent 410 together with the search result of its own node.

  Similar processing is performed in the primary node 440. However, the condition determination means shown in FIG. 7 is embedded in the secondary node 443 under the primary node 440. The condition determination means can specify a disclosure condition for each content. All travel information is classified by content. In this example, information in which “British trip” and “Australia trip” are specified in the content is made public and can be referred from any client terminal. However, the content content “Maldives Travel” is designated as limited release, and the travel information of this content content is only disclosed to the client terminal of the B travel agency with ID 5682. This reflects that the travel plan for Maldives is an exclusive contract with B Travel Agent. The condition determination means determines information to be disclosed based on the client information embedded in the search request, and the secondary node 443 returns only the travel information of “UK travel” and “Australia travel” to the primary node 440. The primary node 440 transmits the search result to the search agent 410 together with the search result of its own node.

  Thus, information is collected in the search agent 410 and the search result is transmitted to the client terminal 400. The reply of the search result from each node to the upper node or from the primary node to the search agent is performed at regular time intervals, and the search result is transmitted to the client terminal 400 at regular time intervals. It is also possible to manually request an update to the latest search result from the client terminal 400.

  Next, it is assumed that a search request is made to the search agent 410 from the client terminal 400 of the B travel agency. In this case, in the secondary node 443, all the information of “British trip”, “Australia trip”, and “Maldive trip” is answered to the primary node 440, and the information is transmitted to the search agent 410 and displayed on the client terminal.

  In order to receive direct access on the Internet as a public server, appropriate security measures are required. In the present invention, only the primary node list server and the primary node list need to be directly connected to the Internet and accessed from the outside, and the lower node list and nodes located on the network inside the enterprise are external. There is no need to receive direct access from. Therefore, it has an excellent structure from the viewpoint of security.

  FIG. 6 is a structural diagram of a table possessed by the ID management means stored in the primary node list server. The client terminal is identified by the ID information 500 and the PASSWORD information 501. If they match, the fact is transmitted to the search agent. If they do not match, a message indicating that authentication cannot be performed is sent to the search agent. The name 502 is an auxiliary data string and is not used directly for searching.

  FIG. 7 shows condition determination means embedded in the node. The content content 600 represents the classification of information. The limited disclosure destination 601 represents the disclosure destination by an ID when limiting the disclosure destination of the information.

(1) Travel industry / primary node: travel planning company / secondary node and below: each department of travel planning company, affiliated company / primary node list: travel industry association / agent: travel agency, etc. Each node is searched by the agent from the travel agency window.
(2) Real estate industry / primary node: construction company / secondary node and below: construction company departments, affiliated companies, management units for events and large-scale condominiums, etc./primary node list: real estate company / agent: real estate company condominium, apartment The information about a house, land, etc. is managed at each node, and each node is searched at a real estate agent window.
(3) Advertising industry / primary node: Broadcasting stations, video distribution sites, newspapers, magazines, etc. (all companies that have the opportunity to distribute advertisements)
・ Secondary node and below, each company department, affiliated company ・ Primary node list: Advertising industry association ・ Agent: Advertising agency The contents and the advertising opportunities associated therewith are managed in each node, and each node is searched from the advertising agency.

It is an outline figure of the present invention and the information retrieval method in a distributed hierarchy base. It is a hardware block diagram of this implementation example. It is a table structure figure of the node information stored in a node list server. FIG. 6 is a flowchart from when a search agent starts a search to when the search ends. It is an example of implementation in the travel industry of the present invention. FIG. 5 is a structural diagram of a table possessed by an ID management means of a primary node list server when identifying a client terminal by ID. FIG. 5 is a structural diagram of a table possessed by a condition determination unit when the disclosure level of information is defined by a node.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Client terminal, 20 ... Search agent, 30 ... Internet line, 40 ... Primary node list server, 41, 42, 43 ... Primary node, 50 ... Local area network, 60 ... Secondary node list server, 61, 62, 63 64 ... secondary node, 70 ... n-order node list server, 71 ... n-order node, 101 ... client reception means, 102 ... client transmission means, 103 ... lower node management means, 104 ... search result collection means, 121 ... node Management means 122 ... ID management means 131 ... search request receiving means 132 ... search execution means 133 ... lower node management means 134 ... search result collection means 135 ... condition determination means 136 ... information DB 141 ... node Management means 151... Search request receiving means 152. Search execution means 153. Management means 154 ... Search result collection means 155 ... Condition determination means 156 ... Information DB, 200 ... Node number, 201 ... URL information, 400, 401 ... Client terminal, 410 ... Search agent, 420 ... Primary node list server, 430, 440 ... Primary node, 431, 441 ... Secondary node list server, 432, 433, 442, 443 ... Secondary node, 450 ... Internet line, 460, 470 ... Local area network, 500 ... ID information, 501 ... PASSWORD Information 502 ... Name 600 ... Content content 603 ... Limited release destination.

Claims (6)

In a distributed hierarchical base composed of a search agent, a primary node list server connected to the search agent through the network, a primary node, an n + 1 order node list server connected to the n order node through the network, and an n + 1 order node,
The search node list only needs to manage the nodes of the corresponding dimension, and is an information search method that can be realized as a very simple mechanism. In a distributed hierarchical base composed of a search agent, a primary node list server connected to the search agent through the network, a primary node, an n + 1 order node list server connected to the n order node through the network, and an n + 1 order node,
An information retrieval method that can be constructed in a relatively small system because each node manages only necessary information and does not manage information such as other departments. In a distributed hierarchical base composed of a search agent, a primary node list server connected to the search agent through the network, a primary node, an n + 1 order node list server connected to the n order node through the network, and an n + 1 order node,
An information search method in which information under any nth-order node can be made independent by making information private to other nodes. In a distributed hierarchical base composed of a search agent, a primary node list server connected to the search agent through the network, a primary node, an n + 1 order node list server connected to the n order node through the network, and an n + 1 order node,
The n + 1st order node must be accessible from one nth order node through the network, and does not depend on whether it can be directly referenced from the agent. Therefore, when the n + 1st order node is installed on the internal network, it is as secure as a public server. An information retrieval method that can be implemented without application. In a distributed hierarchical base composed of a search agent, a primary node list server connected to the search agent through the network, a primary node, an n + 1 order node list server connected to the n order node through the network, and an n + 1 order node,
An information search method that can be implemented very simply by distributing each role and can be realized at a relatively low cost. In a distributed hierarchical base composed of a search agent, a primary node list server connected to the search agent through the network, a primary node, an n + 1 order node list server connected to the n order node through the network, and an n + 1 order node,
An information search method characterized in that traffic on the network is reduced because a search request is not made directly from an agent to all nodes.

Images