| Popis: |
Sentence similarity calculation is a key technology in the field of natural processing and plays an important role in the field of interactive machine translation. This paper combines the important role of prior knowledge in interactive machine translation, and considers that algorithms based on Jaccard algorithm and Wrod2Vec have certain defects in interactive machine translation. Aiming at the phenomenon of a large number of synonyms in the translation field, this paper proposes an improved Word2Vec sentence similarity algorithm based on the synonym table. The algorithm improves the accuracy of sentence similarity calculation by constructing a synonym table and fusing word vectors. The experimental results show that compared with the conventional method, the precision and F-Measure the algorithm in calculating the sentence similarity is higher. |
