Automatic Query Recommendation using ClickThrough Data

Proceedings of the third ACM international conference on Web search and data mining - WSDM '10, 2010

Query recommendations are an integral part of modern search engines. Their goal is to facilitate users' search tasks, as well as help them discover and explore concepts related to their information needs. In this paper, we present a formal treatment of the problem of query recommendation. In our framework we model the user-querying behavior by a probabilistic reformulation graph, or query-flow graph [Boldi et al. CIKM 2008], so that the sequence of queries submitted by a user can be seen as a path on this graph. Assigning score values to queries allows us to define suitable utility functions and to consider the expected utility achieved by performing a random walk on the query-flow graph. Furthermore, providing recommendations can be seen as adding shortcuts in the query-flow graph that "nudge" the reformulation paths of users, in such a way that users are more likely to follow paths with larger expected utility.

2004

Abstract. Search engines are the primary means by which people locate information on the Web. Unfortunately most Web users are not information retrieval experts and there is a tendency for Web queries to be ambiguous and under-specified. Query expansion and recommendation techniques offer one way to solve the ambiguous query problem in Web search, by automatically identifying and adding new terms to a vague query in order to focus the search.

Proceedings of the …, 2009

recently, search engines become more critical for finding information over the World Wide Web where web content growing fast, the user's satisfaction of search engine results is decreased. This paper proposes a method for suggesting a list of queries that are related to the user input query. The related queries are based on previously issued queries by the users. The proposed method is based on clustering process in which groups of semantically similar queries are detected. This facility provides some queries which are related to the queries submitted by users in order direct them toward their required information. This method not only discovered the related queries but also rank them according to a similarity measure. Finally the method has been evaluated using real data sets from the search engine query log.

Int. J. Web Appl., 2013

Query Expansion Methods are proposed to solve many problems of information retrieval systems, but most of these methods do not use the information of interactions between the users and the system. In our approach, we applied a Query Recommendation Algorithm on a list of past user queries, to extract the most associated queries to the input query, and used it in a Probabilistic Query Expansion method, that is constructed as a language model to search in set of candidate terms for the most relevant terms for the initial query that we have to expand. The output of this approach is a table of terms that are candidates for the expansion of the user query, and their values of correlation with the whole query. We did our experiments using the database CISI of the standard collection of test SMART. The results show that the best values are reached by adding fifteen terms to the input queries, using the five most relevant documents of the input query and the five most relevant documents for ...

Query recommendation is considered an effective assistant in enhancing keyword based queries in search engines and Web search software. Conventional approach to query recommendation has been focused on query-term based analysis over the user access logs. In this paper, we argue that utilizing the connectivity of a query-URL bipartite graph to recommend relevant queries can significantly improve the accuracy and effectiveness of the conventional query-term based query recommendation systems. We refer to the Query-URL Bipartite based query reCommendation approach as QUBIC. The QUBIC approach has two unique characteristics. First, instead of operating on the original bipartite graph directly using biclique based approach or graph clustering, we extract an affinity graph of queries from the initial query-URL bipartite graph. The affinity graph consists of only queries as its vertices and its edges are weighted according to a query-URL vector based similarity (distance) measure. By utilizing the query affinity graph, we are able to capture the propagation of similarity from query to query by inducing an implicit topical relatedness between queries. We devise a novel rank mechanism for ordering the related queries based on the merging distances of a hierarchical agglomerative clustering. We compare our proposed ranking algorithm with both naïve ranking that uses the query-URL similarity measure directly, and the single-linkage based ranking method. In addition, we make it possible for users to interactively participate in the query recommendation process, to bridge the gap between the determinacy of actual similarity values and the indeterminacy of users' information needs, allowing the lists of related queries to be changed from user to user and query to query, thus personalizing the query recommendation on demand. The experimental results from two query collections demonstrate the effectiveness and feasibility of our approach.

2013 IEEE 25th International Conference on Tools with Artificial Intelligence, 2013

Commercial web search engines include a query suggestion module so that given a user's keyword query, alternative suggestions are offered and served as a guide to assist the user in formulating queries which capture his/her intended information need in a quick and simple manner. Majority of these modules, however, perform an in-depth analysis of large query logs and thus (i) their suggestions are mostly based on queries frequently posted by users and (ii) their design methodologies cannot be applied to make suggestions on customized search applications for enterprises for which their respective query logs are not large enough or non-existent. To address these design issues, we have developed PQS, a probabilistic query suggestion module. Unlike its counterparts, PQS is not constrained by the existence of query logs, since it solely relies on the availability of user-generated content freely accessible online, such as the Wikipedia.org document collection, and applies simple, yet effective, probabilistic-and information retrieval-based models, i.e., the Multinomial, Bigram Language, and Vector Space Models, to provide useful and diverse query suggestions. Empirical studies conducted using a set of test queries and the feedbacks provided by Mechanical Turk appraisers have verified that PQS makes more useful suggestions than Yahoo! and is almost as good as Google and Bing based on the relatively small difference in performance measures achieved by Google and Bing over PQS.

Second International Conference on the Innovative Computing Technology (INTECH 2012), 2012

Query Expansion Methods are proposed to solve many problems of information retrieval systems, but most of these methods do not use the information of interactions between the users and the system. In our approach, we applied a Query Recommendation Algorithm on a list of past user queries, to extract the most associated queries to the input query, and used it in a Probabilistic Query Expansion method, that is constructed as a language model to search in set of candidate terms for the most relevant terms for the initial query that we have to expand. The output of this approach is a table of terms that are candidates for the expansion of the user query, and their values of correlation with the whole query. We did our experiments using the database CISI of the standard collection of test SMART. The results show that the best values are reached by adding fifteen terms to the input queries, using the five most relevant documents of the input query and the five most relevant documents for each of the recommended queries used, for short and long queries. Our experiments shows also, that for short queries we need to use just the best recommended query in the process of expansion to have a very high value of the Interpolated Average Precision (IAP), but concerning long queries we need to employ also the second best recommended query to have the highest value of IAP. The final experiment was a comparison between our approach and the Rocchio Algorithm, and it shows that our approach gives best results except in the case of using four Recommended Queries (RQs) for short input queries.

—Web recommender systems predict the needs of web users and provide them with recommendations to personalize their pages. Such systems had been expected to have a bright future, especially in ecommerce and E-learning environments. However, although they have been intensively explored in the Web Mining and Machine learning fields, and there have been some commercialized systems, the quality of the recommendation and the user satisfaction of such systems are still not conclusive. In this paper we proposed a more robust approach that leverages search query logs for automatically identifying query groups for a number of different users and record the query logs and their respective sessions. The system uses query reformulation and click graphs which contain useful information on user behavior when searching online. Such information can be used effectively for the task of organizing user search histories into query groups. The proposed technique finds value in combining with keyword semantic similarity and filtering which applies knowledge gained from these query groups in various applications such as providing query suggestions for web personalization by favoring the ranking of search results.