Measuring similarity of Chinese commodity comments in e-commerce websites

摘要

Similarity measurements of comment texts are playing important roles in reducing redundant data of comment texts in e-commerce websites. A novel modeling method is presented in this paper to measure the similarity of comment texts. A simple dictionary is constructed based on high frequency adjective of comment texts. The dictionary is utilized to calculate the common words similarity between comment texts. Finally, similarities of comment texts are gotten by combining common word similarity, HowNet based similarity and length similarity based according to their weights. The experiments show that the proposed method performs better than other methods with the accuracy of 83.33%.