An Approach of Semantic Similarity by Combining HowNet and Cilin
Peiying Zhang, Zhanshan Zhang, Weishan Zhang · 2013
Knowing semantic similarity is very important for many applications of computational linguistics and artificial intelligence. This paper proposes a hybrid approach that combines HowNet and Cilin to calculate semantic similarities in order to address the issue of absence or roughness of words in HowNet. We compare experiment results of our approach with those tests with those produced by Word Net-based similarity measurements. One of the benchmarks is Miller and Charles' list of 30 noun pairs which had been manually designated similarity measurements. We correlate our experiments with those computed by several other methods. Experiments on Chinese word pairs show that our approach is the closet to human similarity judgments.