Personalized Concept-Based Clustering of Search Engine Queries

Abstract

Abstract-A major problem of current Web search is that search queries are usually short and ambiguous, and thus are insufficient for specifying the precise user needs. To alleviate this problem, some search engines suggest terms that are semantically related to the submitted queries so that users can choose from the suggestions the ones that reflect their information needs. In this paper, we introduce an effective approach that captures the user's conceptual preferences in order to provide personalized query suggestions. We achieve this goal with two new strategies. First, we develop online techniques that extract concepts from the web-snippets of the search result returned from a query and use the concepts to identify related queries for that query. Second, we propose a new twophase personalized agglomerative clustering algorithm that is able to generate personalized query clusters. To the best of the authors' knowledge, no previous work has addressed personalization for query suggestions. To evaluate the effectiveness of our technique, a Google middleware was developed for collecting clickthrough data to conduct experimental evaluation. Experimental results show that our approach has better precision and recall than the existing query clustering methods.

Query Click and Text Similarity Graph for Query Suggestions

“…In [13][14][15][16][17][18] authors have used snippet information of clicked URLs or search results returned from a query for query recommendation in different ways. These methods are not general and the extensibility is very low.…”

Efficient Way of User Search Location in Query Processing

“…K.W.-T. Leung, W. Ng [6] suggest that Whenever the user enter the query that forward to the middleware then the query passed to the search engine. Search engine provide result with the some link to the middleware.…”

Building Rich User Search Queries Profiles

“…Some prior work has specifically focused on the personalization of query suggestions and query completion suggestions [3], [5], [8], [13]. All these techniques utilize click through data on documents for estimating user preferences instead of query history-based user profiles for personalization, whose construction we deal with in this paper.…”

“…Prior research has dealt with personalization for search results or building user profiles based on their search results interaction, i.e., clicked urls, content of the visited web pages [3], [4], [8], [13], [18]. These works primarily study users' intents over time for better document ranking [1], [4], [9], [11], [12], [15], [17], [19], [20] or they research query suggestion utilizing document click through [3], [8], [13], [16].…”

“…These works primarily study users' intents over time for better document ranking [1], [4], [9], [11], [12], [15], [17], [19], [20] or they research query suggestion utilizing document click through [3], [8], [13], [16]. Building deep profiles from terse queries poses many interesting challenges: (1) we have a cold-start problem [17], [18], [20], i.e., information about the user needs to be gathered first before queries or results can be personalized; (2) a substantial fraction of queries does not yield to personalization (navigational queries like 'facebook', 'bank of america', etc.)…”