Word Similarity Algorithm Based on WordNet And HowNet
Wu Ling Ren, Jin Ju Guo · Applied Mechanics and Materials · 2012
To make the word similarity calculated results more reasonable and accurate, a new word similarity algorithm is proposed. It uses HowNet primitive hierarchical tree structure, and calculates the two primitives’ distance with the method computing WordNet node distance which considers the tree depth, density, path and connecting intensity, etc. Moreover, algorithm also improves the method that distance into similarity. Finally, this algorithm is compared with related algorithms through experiment. The results show that the proposed algorithm effectively improves the precision and accuracy of word similarity calculation.