Using query contexts in information retrieval

Abstract

User query is an element that specifies an information need, but it is not the only one. Studies in literature have found many contextual factors that strongly influence the interpretation of a query. Recent studies have tried to consider the user's interests by creating a user profile. However, a single profile for a user may not be sufficient for a variety of queries of the user. In this study, we propose to use query-specific contexts instead of user-centric ones, including context around query and context within query. The former specifies the environment of a query such as the domain of interest, while the latter refers to context words within the query, which is particularly useful for the selection of relevant term relations. In this paper, both types of context are integrated in an IR model based on language modeling. Our experiments on several TREC collections show that each of the context factors brings significant improvements in retrieval effectiveness.

Accurate and effective latent concept modeling for ad hoc information retrieval

“…This context can take the form of interest modeling based on historic (or social) behavior, or can be composed of evidences extracted from documents (Finkelstein et al, 2002;White et al, 2009). The latter is better known under the "concept-based retrieval" idiom and received much attention throughout the years (Bai et al, 2007;Bendersky et al, 2011;Chang et al, 2006;Egozi et al, 2011;Metzler, Croft, 2007). The basic idea is to expand the queries with sets of words or multiword terms extracted from feedback documents.…”

Contextual Search: A Computational Framework

“…Azzopardi [12] gives a thorough study that starts from theoretical issues, investigates whether and how language models can be an efficient and effective theoretical framework for contextual search, and ends with experiments. Bai et al [14,15] are examples of text window-based context papers with co-occurrence analysis, an interesting modeling of contextual factors based on language models and an analysis of domain knowledge and language model combination. Bartholomew et al [17] provide a perspective of the factorial models that are relevant to the notion of computational framework used in this survey.…”

“…Anastácio et al [8] address semianonymity when geographical variables are exploited. Bai et al [14,15] use TREC collections. Bian et al [23] use LETOR and TREC collections.…”

Integrating Query Context and User Context in an Information Retrieval Model Based on Expanded Language Modeling

“…θ 0 q can be created by MLE (Maximum Likelihood Estimation), as in [7]. We will describe the details to construct θ s q and θ u q in the following sections.…”