Determining the similarity of short text snippets, such as search queries, works poorly with traditional document similarity measures (e.g., cosine), since there are often few, if any, terms in common between two short text snippets. We address this problem by introducing a novel method for measuring the similarity between short text snippets (even those without any overlapping terms) by leveraging web search results to provide greater context for the short texts. In this paper, we define such a similarity kernel function, mathematically analyze some of its properties, and provide examples of its efficacy. We also show the use of this kernel function in a large-scale system for suggesting related queries to search engine users.
Description
A web-based kernel function for measuring the similarity of short text snippets
%0 Conference Paper
%1 Sahami06querySimilarity
%A Sahami, Mehran
%A Heilman, Timothy D.
%B WWW '06: Proceedings of the 15th international conference on World Wide Web
%C New York, NY, USA
%D 2006
%I ACM
%K 06 Sahami query search similarity term
%P 377--386
%R http://doi.acm.org/10.1145/1135777.1135834
%T A web-based kernel function for measuring the similarity of short text snippets
%U http://portal.acm.org/citation.cfm?id=1135834
%X Determining the similarity of short text snippets, such as search queries, works poorly with traditional document similarity measures (e.g., cosine), since there are often few, if any, terms in common between two short text snippets. We address this problem by introducing a novel method for measuring the similarity between short text snippets (even those without any overlapping terms) by leveraging web search results to provide greater context for the short texts. In this paper, we define such a similarity kernel function, mathematically analyze some of its properties, and provide examples of its efficacy. We also show the use of this kernel function in a large-scale system for suggesting related queries to search engine users.
%@ 1-59593-323-9
@inproceedings{Sahami06querySimilarity,
abstract = {Determining the similarity of short text snippets, such as search queries, works poorly with traditional document similarity measures (e.g., cosine), since there are often few, if any, terms in common between two short text snippets. We address this problem by introducing a novel method for measuring the similarity between short text snippets (even those without any overlapping terms) by leveraging web search results to provide greater context for the short texts. In this paper, we define such a similarity kernel function, mathematically analyze some of its properties, and provide examples of its efficacy. We also show the use of this kernel function in a large-scale system for suggesting related queries to search engine users.},
added-at = {2010-01-25T21:19:40.000+0100},
address = {New York, NY, USA},
author = {Sahami, Mehran and Heilman, Timothy D.},
biburl = {https://www.bibsonomy.org/bibtex/26f6d9db4ff5494cbedb410a15dc8ba4f/lee_peck},
booktitle = {WWW '06: Proceedings of the 15th international conference on World Wide Web},
description = {A web-based kernel function for measuring the similarity of short text snippets},
doi = {http://doi.acm.org/10.1145/1135777.1135834},
interhash = {ed02f66d2bf7e17397c9b7de4c90e2dc},
intrahash = {6f6d9db4ff5494cbedb410a15dc8ba4f},
isbn = {1-59593-323-9},
keywords = {06 Sahami query search similarity term},
location = {Edinburgh, Scotland},
pages = {377--386},
publisher = {ACM},
timestamp = {2010-01-25T21:19:40.000+0100},
title = {A web-based kernel function for measuring the similarity of short text snippets},
url = {http://portal.acm.org/citation.cfm?id=1135834},
year = 2006
}