Kernelized Hashcode Representations for Relation Extraction
This work addresses computational bottlenecks for researchers and practitioners in NLP, particularly in biomedical relation extraction, by enabling faster and more accurate classification with general methods, though it builds incrementally on existing KLSH techniques.
The paper tackled the scalability problem of kernel methods in relation extraction by proposing random subspaces of kernelized locality-sensitive hashing (KLSH) codes for efficient explicit representations, resulting in significant accuracy improvements and orders-of-magnitude speedup compared to state-of-the-art kernel methods.
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.