A new patent application from Microsoft considers ways to present search results to searchers in clusters with meaningful names. This kind of results clustering is sometimes referred to as domain clustering because it normally involves grouping together search results from the same domain.
Published on February 2, 2006, the results clustering patent application was originally filed on July 13, 2004, and is assigned to Microsoft Corporation.
Query-based snippet clustering for search result grouping
Inventors: Hua-Jun Zeng, Qicai He, Guimei Liu, Zheng Chen, Benyu Zhang, and Wei-Ying Ma
US Patent Application 20060026152
A clustering architecture that dynamically groups the search result documents into clusters labeled by phrases extracted from the search result snippets. Documents related to the same topic usually share a common vocabulary. The words are first clustered based on their co-occurrences, and each cluster forms a potentially interesting topic. Keywords are chosen and then clustered by counting co-occurrences of pairs of keywords. Documents are assigned to relevant topics based on the feature vectors of the clusters.
This results clustering process is intended to make it easier for searchers to find what they are looking for by clustering results into groups and providing a user with the ability to quickly determine whether different clusters match what they are looking for. Besides, clusters that may be more relevant for a query used by a searcher are ranked higher than ones that aren’t.
Some search engines display groupings with the results that they show searchers, but those differences in those clusters are based on content similarity.
From an analysis of the results originally returned upon a query, these groupings are created dynamically, looking at the common language used in titles and snippets. Interestingly, the results clustering patent application notes that words and phrases from titles will likely be given more weight than from snippets.