A New Approach to plagiarism Detection Using Cellular Learning Automatons and Semantic Role Labeling

Abstract

Plagiarism is removal and to put it in their own name the ideas or words of others. With the Increasing progress of the Internet and the proliferation of online articles, scientific theft has also become easier. Many systems have been developed today to detect plagiarism. Most of these systems are based on lexical structure and string matching algorithms. Therefore, these systems can hardly detect recovery robberies, placement of synonyms. This paper presents a method for identifying plagiarism based on semantic role labeling and cellular learning automata. In this paper, cellular learning automata are used to locate the processed words. Semantic role labeling specifies the role of words in sentence. Comparison operations are performed for all sentences of the original text and suspicious text. Results of the experiments on PAN-PC-11 corpus demonstrate the proposed method improves values of evaluation parameters such as recall, precision and F-measure, comparing to previous approaches in plagiarism detection.

Keywords

• plagiarism detection
• cellular learning automata
• semantic role labeling
• semantic similarity