Two-Stage Hashing for Fast Document Retrieval

This work fulfills sublinear time Near-est Neighbor Search (NNS) in massive-scale document collections. The primary contribution is to propose a two-stage unsupervised hashing framework which harmoniously integrates two state-of-the-art hashing algorithms Locality Sensitive Hashing (LSH) and Iterative Quantization (ITQ). LSH accounts for neighbor candi-date pruning, while ITQ provides an ef-ficient and effective reranking over the neighbor pool captured by LSH. Further-more, the proposed hashing framework capitalizes on both term and topic similar-ity among documents, leading to precise document retrieval. The experimental re-sults convincingly show that our hashing based document retrieval approach well approximates the conventional Informa-tion Retrieval (IR) method in terms of re-trieving semantically similar documents, and meanwhile achieves a speedup of over one order of magnitude in query time.