Abstract Information retrieval involves finding some required information in a collection of
information or in databases. The information or database need not necessarily be in one …

5,222,236 A 6, 1993 Potash et al................. 707/102 5,257,577 A 1 1/1993
Clark.......................... 100.99 5,594,660 A 1/1997 Sung et al................ T15, 500.1 5,606,609 A …

5,606.609 A 2f1997 Houser et al. with each of the documents on the network. The click
distance 5,848.404 A 12, 1998 Hafner et al................... 707/3 is measurement of the number …

A field-weighted search combines statistical information for each term across document
fields in a suitably weighted fash ion. Both field-specific term frequencies and field and docu …

A neural network is used to process a set of ranking features in order to determine the
relevancy ranking for a set of docu ments or other items. The neural network calculates a pre …

In recent years, semantic search for relevant documents on web has been an important topic
of research. Many semantic web search engines have been developed like Ontolook …

US PATENT DOCUMENTS Methods of providing a document relevance score to a docu
5.222. 236 A 6, 1993 Potash et al. ment on a network are disclosed. Computer readable …

EP O950961 A2 10, 1999 EP O950961 A3 10, 1999 EP 1050830 A2 11/2000 EP 1120717
A2 8, 2001 EP 1282060 A2 2, 2003 EP 155777O A1 7/2005 JP 1009 1638 4f1998 JP …

US PATENT DOCUMENTS component having a ranking component that can be used to
rank search results as part of a query response. In one embodi ment, the ranking component …

Search results of a search query on a netWork are ranked according to an additional ranking
function for the prior prob ability of relevance of a document based on document prop erty …