JP2007213601A - Method for searching a plurality of databases and method for searching for literature between the plurality of databases - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for literature search using a plurality of databases used from at least one server by use of at least one search engine. <P>SOLUTION: The number of records is determined and is reported for each the database, and frequency of a hit or appearance of searched query term is determined and is reported together with identification of the record of the database corresponding to the hit. The reports from the plurality of databases are given to a user terminal, i.e., a client, and client software calculates a related score for each the record on the basis of the number of the records inside the database, the number of the records each having at least one hit, and the number of the hits about each the record. By local calculation from unified data, coherent ranking is achieved as if resulting from a single database, for all the literatures. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to literature searching and searching, and more particularly to those via a network.

  For over 20 years, information services have provided access to multiple databases. For example, Dialog Information Services, now known as Knight-Ridder Information, Inc., has hundreds of databases available to searchers ( (Also known as a collection). Some of these databases contain bibliographic summaries, while others contain full-text literature. A searcher can apply a query to one or more databases. Initially, the searcher selects an individual database of interest based on past experience, or selects a set of databases selected by an information provider and associated with a particular topic. For example, a searcher may select the topic of patents. Information services on this topic have compiled several databases limited to patents. When a query is applied to a group of databases, the information service retrieves the number of hits in each database. The searcher then accesses the database of interest and browses the individual records. The system was originally designed for librarians and professional researchers who knew where to look for the information they wanted.

  As wide area networks such as the Internet become available, not only search professionals but also amateur users have new opportunities in search. A new type of information provider has emerged that uses public databases as well as private ones to provide bibliographic research data and literature to users. If a user is interested in a certain topic, such as a patent, the user may not know what resources can be gathered for searching and the location of those resources. Since resources are often changed, the searcher will have less interest in the source of the response compared to the appropriateness of the response. It is recognized that distributed collections available over a wide area network can be treated as a single collection. Each sub-collection is searched individually and the reports are combined into a single list. It is also known that the literature can be ranked and weighted by a search engine according to an algorithm, taking into account the properties of a particular collection. The literature scores can be normalized to obtain a score that would be obtained if the individual literature collections were merged into a single consolidated collection.

  One of the problems existing in the prior art is that the score for each document is not absolute and depends on the statistics in each collection and the algorithm associated with the search engine. A second problem that exists is that the standard prior art procedure requires two passes. In the first pass, statistics are gathered from each search engine to calculate the weights for each query term. In the second step, the information from the first step is sent back to each search engine, which then assigns a particular weight or score to each hit or identified document. A third problem that exists is that the prior art requires that all collections use the same search engine.

  The purpose of this invention is to rank documents based on consistent criteria so that if the same document appears in two different databases, it will be scored in the same way when the results are merged. To devise a method for searching multiple collections in a single pass, with attachment. It is not necessary to use the same search engine for all collections.

Summary of the Invention

  The above objective is accomplished with a document search and search method that requires each participating search engine server to return statistics about the terms of each query in each of the returned documents. The final relevance score is then calculated at the client terminal, not the server. In this manner, all relevance scores are processed at the client in the same manner regardless of search engine differences.

Best mode for carrying out the invention

Referring to FIG. 1, the query indicated by query block 11 is clarified by the user and provided to the terminal or client system. This query is transmitted electronically to the network interface 13. A network interface is an information service that has access to a source 17 that has a database associated with the subject of the query. These databases deployed on other servers are polled simultaneously via the communication channel 15. Communication channel 15 may be a wide area network link to source 17. The Internet is one model for such wide area network links and remote sources. The query is given to the search engine. These search engines are represented in columns 20, 30, and 40, with each search engine accessing an associated database at block 19. Each search engine may have its own operational characteristics, such as Boolean logic, statistical inference. Each database generates a report containing the number N of records in the database, indicated by block 21. The report also includes the number of times each search term appears in the literature that responds to the query. This quantity, N _TERM, is indicated by block 23. Third, the report generates a document identification number for each document that contains a hit along with the number of times each search word appears, as indicated by block 25. From this information, as shown in block 27, the score for each document is calculated by the client software using its own algorithm. For example, the formula for calculating the score is:

Where t _f = the number of times the word appears in the document, idf = log (N / N _TERM ), where N and N _TERM are the sum of N and N _TERM values reported by all collections It is. The calculated score is sent to the output buffer indicated by block 29, which shifts the top M scores requested by the querying person from calculation block 27. Note that there is only one pass through this database. The calculated score is treated as an absolute value.

  In alternative embodiments, optional parameters may be reported for use in the algorithm. Block 26 indicates that the frequency of the most frequent word in each document is reported for the purpose of using the client's more sophisticated ranking formula. Another optional data reduction step is that each search engine may calculate a score for document relevance in a manner known in the prior art. From this data, the search engine will be able to pre-select the top M hits in the database. Here, M is the maximum number of hits requested by the user.

As an example, a search query may involve documents with the word “graphical user interface”. Table 1 below shows a report generated by a search engine that selected some of the highest ranking documents. This report is returned to the user's client software, where the user applies an algorithm such as that in formula (1) above using the word frequency data and literature frequency data returned by each search engine. Therefore, a local calculation of the document weight is performed for each query word, and N _TERM and N (= the number of documents) returned from each collection are combined. Thus, the word weights are exactly the same as if these collections were a single collection. Scoring is completely consistent even when different search engines participate in the search, and the same document appearing in two different collections will always receive the same score.

It is a block diagram of the system in the present invention.

Explanation of symbols

11 Query 13 Network interface 15 Communication channel 17 Source 19 Database 20 Search engine 30 Search engine 40 Search engine

Claims (9)

A method for searching a plurality of databases that are distributed and accessible to clients via one or more search servers, comprising:
(A) providing a search query from a client to each database and each associated server;
(B) obtaining statistics for each database at each client from each server;
(C) obtaining information about a document obtained as a result of applying a query to a database at each client from each server;
(D) at the client, calculating a score for each document using the statistics and the information, the calculated scores for all databases as if they were combined as a single database To search multiple databases that appear as applicable.   The method of claim 1, wherein the statistics for the collection include the size of the collection in relation to the number of records.   The method of claim 1, wherein the information about each document includes the number of times each search term appears in the document.   The method of claim 1, wherein the information about each database includes the number of documents that contain each search term. A method for searching a plurality of databases that are distributed and accessible to clients via one or more servers, comprising:
(A) accessing each database from a client;
(B) providing a search query from a client to a server associated with each database;
(C) obtaining statistics for each database at the client;
(D) obtaining statistical information about relevant documents obtained as a result of applying the query to the database at the client;
(E) calculating a score for a related document using the statistics and the information at the client, wherein the score calculated for a document is independent of the database in which it appears; A method for searching multiple databases. A method of searching for documents among a plurality of databases in response to a search query,
(A) providing a search query for each database;
(B) determining the number of records for each database;
(C) applying a search query to each of the databases to record the number of hits for each search query word and the identification of a database record having at least one hit in the number of hits;
(D) counting records having at least one hit for each of the databases and for each query term;
(E) reporting the associated score of each record for the search query to the user, calculated using the results of steps (b), (c) and (d), between a plurality of databases How to search for literature.   7. The method of claim 6, further defined by selecting several databases and having more than one search engine for the databases prior to providing the same search query for all databases.   8. The method of claim 7, wherein by selecting a number of records to be reviewed between the number of databases, the number of records having the highest relevance score for a search query. .   9. The method of claim 8, further defined by pre-selecting a number of records prior to calculating an associated score.

Images