摘要 : Sentence similarity is used in various fields, such as the mining of text, information retrieval from the web, and dialogue-based system. This research mainly focuses on calculating the sentence-length similarity between very brie... 展开 Sentence similarity is used in various fields, such as the mining of text, information retrieval from the web, and dialogue-based system. This research mainly focuses on calculating the sentence-length similarity between very brief texts. It provides a method that works on the implicit word order and contextual relations in the phrases. A combination of data from the corpus statistics and hierarchical database is used to determine the computation of similarity between sentence pairs. Our technique can simulate human sensible knowledge according to the usage of a lexical database, and it may be applied to other areas according to the incorporation of corpora statistics. Numerous applications that involve the representation and finding of text knowledge can make use of the suggested approach. Studies done on two sets of chosen sentence pairings reveal that the suggested approach offers a similarity metric that significantly correlates with human intuition. 收起