Context-dependent conceptualization

Conceptualization seeks to map a short text (i.e., a word or a phrase) to a set of concepts as a mecha-nism of understanding text. Most of prior research in conceptualization uses human-crafted knowl-edge bases that map instances to concepts. Such approaches to conceptualization have the limitation that the mappings are not context sensitive. To overcome this limitation, we propose a framework in which we harness the power of a probabilis-tic topic model which inherently captures the se-mantic relations between words. By combining la-tent Dirichlet allocation, a widely used topic model with Probase, a large-scale probabilistic knowledge base, we develop a corpus-based framework for context-dependent conceptualization. Through this simple but powerful framework, we improve con-ceptualization and enable a wide range of applica-tions that rely on semantic understanding of short texts, including frame element prediction, word similarity in context, ad-query similarity, and query similarity.