Browse Prior Art Database

Method and System for Increasing the Relevancy of Search Results Using Categories of Related Terms

IP.com Disclosure Number: IPCOM000204188D
Publication Date: 2011-Feb-17
Document File: 4 page(s) / 122K

Publishing Venue

The IP.com Prior Art Database

Related People

Jagannatha Narayanareddy: INVENTOR [+3]

Abstract

A method and system for increasing the relevancy of search results using categories of related terms is disclosed.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Method and System for Increasing the Relevancy of Search Results Using Categories of Related Terms

Abstract

A method and system for increasing the relevancy of search results using categories of related terms is disclosed. 

Description

Disclosed is a method and system for increasing the relevancy of search results using categories of related terms.

Currently, many search-base applications use indexed data to query a keyword and provide best-related results.  The indexed data is in the form of documents.  The content in the documents may be sufficient or insufficient.  If the documents contain insufficient content, differentiating two small content documents mathematically becomes difficult.  Also, if the documents contain insufficient content, the probability of getting irrelevant results is higher even when using a lexicographical searcher.  Therefore, search-based applications are not able to provide relevant results.

An application that provides related terms corresponding to a searched query is one example, which uses insufficient content in its documents.  Here, a database of terms that has been categorized in a hierarchical structure is used.  An exemplary hierarchical structure is as shown in Table 1.  Here, Ri represents the stored terms and CATi represents the category hierarchy corresponding to the term.  A document Di consists of a term Ri and the corresponding category CATi.

Document No.

Related Terms

Category

D1

R1

CAT1: C1/C2/C3

D2

R2

CAT2: C4/C5

D3

R3

CAT3: C1/C2

D4

R4

CAT4: C4

D5

R5

CAT5: C6

D6

R6

CAT6: C6/C7

D7

R7

CAT7: C4/C5

D8

R8

CAT8: C1/C2/C8

Table 1

The method and system disclosed herein uses the occurrence of category hierarchy in search related terms to refine the search results in order to increase the relevancy of the search results.  Here, the main motive is to provide relevant results according to the user’s intention.  To achieve this, the disclosed method uses the bucket system to determine high-density nodes in the category hierarchy tree.

Here, the disclosed method increases the relevancy of search results using the occurrence of category hierarchy.  The disclosed method involves computation of occurrence of category hierarchy in the search results.  Thereafter, a new score for the search related terms is computed by providing a higher preference to a higher density category node lower in depth in the category hierarchy along with a “tf-idf” score as explained later.  Subsequently, the searched related terms are re-ordered based on the new score in decreasing order.  The complete process of indexing and searching in accordance with the disclosed method is illustrated in Fig. 1.

Figure 1

To compute the occurrence of the category hierarchy, the top m results along with its parental relation in tree form of categories of search results is cumulated.  In other words, the top m search results from a lexicographical searcher for a particular query are considered and the occurrence of each categ...