2017/11/10 by Sahil Garg, Garg, Sahil, Aram Galstyan +9
Biochemistry, Genetics and Molecular Biology · Computer Science · Mathematics · #Advanced Image and Video Retrieval Techniques #Artificial intelligence #Class (philosophy) #Computation and Language (cs.CL) #Computer science #Data mining #FOS: Computer and information sciences #Hash function #Hash table #Information Retrieval (cs.IR) #Kernel (algebra) #Kernel method #Locality #Locality-sensitive hashing #Machine Learning (cs.LG) #Machine Learning in Bioinformatics #Machine learning #Mathematics #Pattern recognition (psychology) #Relation (database) #Relationship extraction #Representation (politics) #Scalability #Speedup #Support vector machine #Text and Document Classification Technologies #cs.CL #cs.IR #cs.LG
paper · pdf · doi:10.48550/arxiv.1711.04044
To appear in the proceedings of conference, AAAI-19
openalex publication_date 2017/11/10 · arxiv created 2019/05/20 · arxiv updated 2019/05/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
Kernel methods have produced state-of-the-art results for a number of NLP tasks such as relation extraction, but suffer from poor scalability due to the high cost of computing kernel similarities between natural language structures. A recently proposed technique, kernelized locality-sensitive hashing (KLSH), can significantly reduce the computational cost, but is only applicable to classifiers operating on kNN graphs. Here we propose to use random subspaces of KLSH codes for efficiently constructing an explicit representation of NLP structures suitable for general classification methods. Further, we propose an approach for optimizing the KLSH model for classification problems by maximizing an approximation of mutual information between the KLSH codes (feature vectors) and the class labels. We evaluate the proposed approach on biomedical relation extraction datasets, and observe significant and robust improvements in accuracy w.r.t. state-of-the-art classifiers, along with drastic (orders-of-magnitude) speedup compared to conventional kernel methods.