Measuring Similarity between Sentence Fragments

Sentence fragment has a wide range of applications, such as short text mining, flow diagram search based on label similarity and so on. Existing methods aren´t entirely appropriate for measuring similarity between sentence fragments since they were originally designed for complete sentences or long texts. So we pay more attention to proper nouns which carry important information in sentence fragments. We then propose a novel measuring method applicable for sentence fragments or even short sentences. It calculates the similarity based on the edit distance model instead of traditional vector space model. Besides, manual weight factors are introduced in order to meet the needs of different situations. Our experiments demonstrate that our method outperforms existing methods.