Method of searching documents and a service for searching documents

• US 7,693,910 B2
• Filed: 02/15/2002
• Issued: 04/06/2010
• Est. Priority Date: 11/19/1998
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Iwayama, Makoto, Nishioka, Shingo, Takano, Akihiko, Niwa, Yoshiki
• Primary Examiner(s): Breene; John E
• Assistant Examiner(s): Kerzhner; Aleksandr
• Application Number:
  US10/075,283
• Publication Number:
  US 20020078030A1
• Time in Patent Office: 2,972 Days
• Field of Search: 707 1- 10, 715/500, 715/513
• US Class Current: 707/766
• CPC Class Codes: G06F 16/30 of unstructured textual data document management systems G06F16/93
• CPC Class Codes: G06F 16/345 Summarisation for human users
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

This invention aims to provide an efficient means of performing a document search wherein relevance between plural document databases is examined. A summary making module and a search module are provided to document databases, and these are connected to a network as a server. A client obtains a relevant set of documents in a specified document database via this summary from a set of documents in the specified document database. The summary obtained is sent to another server, and a search is performed according to the summary in a document database in the server to which the summary is transferred.

Claims

1. A computer implemented service of searching documents wherein servers comprising document databases and programs to manipulate said databases are dispersed over a network and a client connected to said servers performs a document search, said service providing a document search method comprising the steps of:

making a first search input of a set of keywords, fragments of a document or any desired set of documents to a first document database;

conducting a first search of said first document database based on said first search input;

retrieving at least one document from said first document database as a result of said first search;

selecting and inputting at least one of said at least one retrieved document to said first document database;

making a weighted term list, whereby each term is given a first weight, from said at least one inputted document, the first weight of each term reflecting the importance of the term in the first document database;

wherein a user interface is provided in which any of said at least one document retrieved by said first search is selected or deselected, and a set of said at least one document that is selected via said interface is used in making the first weighted term list; and

selecting at least one term from said weighted term list,

conducting a second search of a second document database which differs from said first document database using said at least one term selected from said weighted term list as a second search input for performing said second search;

retrieving at least one document from said second document database as a result of said second search,

calculating a second weight of each term used as the second search input, which reflects the importance of each said term, to each document retrieved from the second document database by the second search;

finding out a relevance of each document retrieved from the second document database by calculating an overall weight on both said weighted term list and the calculated second weights for each document with respect to terms common in both; and

displaying results of said finding step on a display unit.

2. The computer implemented service of searching documents as defined in claim 1, wherein the client transmits the set of said at least one document selected via said interface to a first server where said first document database is stored, receives the weighted term list from said first server, sends the second search input to a second server where said second document database is stored, and receives the result from the second search of said second document database.

3. The computer implemented service of searching documents as defined in claim 2, wherein the client sends the second search input corresponding to said summary reflecting a user'"'"'s evaluation of the summary to a second server where said second document database is stored;

the second server calculates the second weight of each term;

the second server finds out the relevance of each document derived from the second document database; and

the client receives the search result from the search of said second document database in accordance with the relevance.

4. A computer implemented service of searching documents wherein servers comprising document databases and programs to manipulate said databases are dispersed over a network and a client connected to said servers performs a document search, said service providing a document search method comprising the steps of:

making a first search input of a set of keywords, fragments of a document or any desired set of documents to a first document database;

conducting a first search of said first document database based on said first search input;

retrieving at least one document as a result of said first search;

selecting and inputting at least one of said at least one retrieved document to said first document database;

making a weighted term list, whereby each term is given a first weight, from said at least one inputted document, the first weight of each term reflecting the importance of the term in the first document database;

selecting at least one term from said weighted term list;

conducting a second search of a second document database which differs from said first document database using said at least one term selected from said weighted term list as a second search input for performing said second search;

wherein the client transmits said at least one document to a first server where said first document database is stored, receives the weighted term list from said first server, sends the second search input to a second server where said second document database is stored, receives the result of the second search of said second document database, and calculates a second weight of each term used as the second search input, which reflects the importance of each said term to each document in the result of the second search;

using said weighted term list and the calculated second weights to determine a relevance of each document in the result of the second search; and

displaying results of said using step on a display unit.

5. The computer implemented service of searching documents as defined in claim 4, wherein the client has a user interface in which any of said at least one document retrieved by said first search is selected or deselected, and in which document databases are selected to be searched, a set of said at least one document thus selected is sent to the first server as specified by the user, the weighted term list of the set of documents is received from said first server, and the received weighted term list is sent to the second server comprising said second document database.