Measuring Semantic Similarity of Bengali Texts with Parts-of-Speech Tags and Word-Level Semantics

摘要: The semantic textual similarity is essential for many applications related to natural language processing. But measuring the not an easy task. Because there are different types of sentences and diversities sentences’ structure make assessing a formidable When two texts lexicographically dissimilar but semantically similar, traditional lexical matching cannot return actual degree similarity. Besides these, lack well-recognized processing resources Bengali makes calculation difficult In this paper, we tried measure using word-level parts-of-speech tags. To assess similarity, exploit tagger pre-trained word-embedding model. Then, maximum word-to-word words employed if belong identical tag. We also introduced grammatical role level in our proposed method validate performance method, conducted experiments on publicly available benchmark dataset. results demonstrated that effective achieved state-of-the-art performance.