Dual-Space Re-ranking Model for Document Retrieval
WADE, VINCENT PATRICK
Metadata:Show full item record
The field of information retrieval still strives to develop models which allow semantic information to be integrated in the ranking process to improve perform- ance in comparison to standard bag-of- words based models. A conceptual model has been adopted in general- purpose retrieval which can comprise a range of concepts, including linguistic terms, latent concepts and explicit knowledge concepts. One of the draw- backs of this model is that the computa- tional cost is significant and often in- tractable in modern test collections. Therefore, approaches utilising concept- based models for re-ranking initial re- trieval results have attracted a consider- able amount of study. This method en- joys the benefits of reduced document corpora for semantic space construction and improved ranking results. However, fitting such a model to a smaller collec- tion is less meaningful than fitting it into the whole corpus. This paper proposes a dual-space model which incorporates external knowledge to enhance the space produced by the latent concept method. This model is intended to produce global consistency across the semantic space: similar entries are likely to have the same re-ranking scores with respect to the latent and manifest concepts. To illustrate the effectiveness of the pro- posed method, experiments were con- ducted using test collections across dif- ferent languages. The results demon- strate that the method can comfortably achieve improvements in retrieval per- formance.
Other Titles:In the Proceedings of the 23rd International Conference on Computational Linguistics, COLING 2010