| 摘要 |
An information searching apparatus and method capable of producing good search results even in the case where specified search keywords are found separately on a plurality of documents. Upon receipt of a search request from a client, a first search process searches an index memory to find tentative search results. According to predetermined rules, it is determined to which group each tentative search result should belong. A second search process then searches documents at an upper hierarchical level in the same group, thus finding keyword hits in an upper-level document. A higher priority is then given to a group of documents in which the second search process has found keyword hits. Finally a list of locators is compiled and sent back to the requesting client to indicate the locations of relevant documents, where the locators are arranged in descending order of the display priority of corresponding groups.
|
