An NLP based plagiarism detection approach for short sentences

Abstract

The notable issue in the fields of plagiarism detection is, to assess the semantic similarity between obfuscated sentences, and it becomes more completed in case of short sentences (only 4-8 words). An innovative approach, typed dependencies relationship (TDR), based on Natural Language processing is presented for detecting plagiarism on short sentences. In this study proposed approach performed on previous datasets of short sentences and compared results with 3 state-of-art methods. The investigation shows that the proposed calculation has exceptional execution in taking care of sentences with complex linguistic structure.