Search processing with automatic categorization of queries

• US 20060122979A1
• Filed: 12/06/2004
• Published: 06/08/2006
• Est. Priority Date: 12/06/2004
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Joshi, Deepa, Parikh, Jignashu, Kapur, Shyam
• Granted Patent: US 7,620,628 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3326 using relevance feedback from the user, e.g. relevance feedback on documents, documents sets, document terms or passages
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99936 Pattern matching access

Abstract

Search results are processed using search requests, including analyzing received queries in order to provide a more sophisticated understanding of the information being sought. A concept network is generated from a set of queries by parsing the queries into units and defining various relationships between the units. From these concept networks, queries can be automatically categorized into categories, or more generally, can be associated with one or more nodes of a taxonomy. The categorization can be used to alter the search results or the presentation of the results to the user. As an example of alterations of search results or presentation, the presentation might include a list of “suggestions” for related search query terms. As other examples, the corpus searched might vary depending on the category or the ordering or selection of the results to present to the user might vary depending on the category. Categorization might be done using a learned set of query-node pairs where a pair maps a particular query to a particular node in the taxonomy. The learned set might be initialized from a manual indication of which queries go with which nodes and enhanced has more searches are performed. One method of enhancement involves tracking post-query click activity to identify how a category estimate of a query might have varied from an actual category for the query as evidenced by the category of the post-query click activity, e.g., a particular hits of the search results that the user selected following the query. Another method involved determining relationships between units in the form of clusters and using clustering to modify the query-node pairs.

Claims

1. A computer-implemented method for associating user search queries with nodes of a taxonomy, wherein each node of the taxonomy represents a category into which queries can be categorized, the method comprising:

accepting a user search query;

generating a set of one or more query units corresponding to concepts represented in the user search query using a concept network that relates units to other units;

accessing a taxonomy data structure, wherein the taxonomy is hierarchical with each node representing a category and a plurality of nodes having associated therewith as node-specific concept networks;

testing the set of one or more query units against a set of unit-node pairs to determine if the set includes one or more recorded units matching the set of one or more query units;

if the set of one or more query units does not match a unit-node pair, testing a set of one or more query units against similar units in the set of unit-node pairs; and

for at least a plurality of user search queries, when user action is taken to independently indicate a category for the user search queries, modifying the set of unit-node pairs to reflect the categorization of the user search queries.

5. The method of claim 1, further comprising modifying search results for the user search query according to a categorization of that user search query.

6. The method of claim 1, further comprising modifying a document corpus searched according to a categorization of that user search query.

7. The method of claim 1, further comprising annotating the user search query according to a categorization of that user search query.

8. The method of claim 1, further comprising generating a list of related search terms according to a categorization of the user search query.

9. The method of claim 1, comprising populating unit-node pairs using clusters, wherein units that are members of a given cluster are more likely to be paired with a node with which other cluster members units are associated due to their clustering.

10. The method of claim 1, comprising populating unit-node pairs using unit frequencies on a vertical-by-vertical basis.

2. The method including 1, further comprising reading in an initial set of unit-node pairs into storage for the set of unit-node pairs, wherein the initial set is manually generated and generated prior to the step of testing.

3. The method including 1, wherein the user action to indicate a category for a user search query comprises:

presenting a set of categorized search results in response to the user search query;

recording an indication of the hit within the categorized search results that a user selects; and

associating, using a unit-node pair, the units derived from the user search query with the category associated with the user-selected hit.

4. The method of claim 3, wherein the set of categorized search results is search results derived from a search of a structured topical directory of web pages.