Arabic-English Cross-language Plagiarism Detection using Winnowing Algorithm

摘要

The availability of information in electronic forms and the availability of automatic translation machines has led to increased cross-language plagiarism. Manual detection of cross-language plagiarism is difficult, as such, developing an automatic system to detect such plagiarism is necessary. Although, there are a number of studies on detecting cross-language plagiarism in the form of Euro-English, Malay-English and Indonesian-English, there remains few studies concerned with the detection of Arabic-English cross-language plagiarism. This study proposes an Arabic-English cross-language plagiarism detection tool using the Winnowing algorithm. We evaluate its performance in terms of precision and recall on a data set consisting of Wikipedia articles. The performance of the proposed tool proved good with 97% precision, 81% recall and 89% F-measure evaluation metrics. The results show that the Winnowing algorithm can be used effectively to detect Arabic-English cross-language plagiarism.