A major difficulty in applying word vector embeddings in information retrieval is in devising an effective and efficient strategy for obtaining representations of compound units of text, such as whole documents, (in comparison to the atomic words), for the purpose of indexing and scoring documents. Instead of striving for a suitable method to obtain a single vector representation of a large document of text, we aim to develop a similarity metric that makes use of the similarities between the individual embedded word vectors in a document and a query. More specifically, we represent a document and a query as sets of word vectors, and use a standard notion of similarity measure between these sets, computed as a function of the similarities between each constituent word pair from these sets. We then make use of this similarity measure in combination with standard information retrieval based similarities for document ranking. The results of our initial experimental investigations show that our proposed method improves MAP by up to 5.77%, in comparison to standard text-based language model similarity, on the TREC 6, 7, 8 and Robust ad-hoc test collections.
CCS Concepts•Information systems → Content analysis and feature selection;