A Platform Framework for Cross-Lingual Text Relatedness Evaluation and Plagiarism Detection

Research work related to plagiarism detection methods in dealing with monolingual texts (e.g. English texts) have been well established in recent years. However, little attention has been paid to facilitate plagiarism detection in cross-lingual text collections (e.g. English and Chinese texts). In this paper we present a system platform to evaluating text similarity and relatedness in multilingual text collections for plagiarism detection. First, we utilized a number of selected texts in Chinese-English parallel corpora collected from Internet to train text classifiers based on the Support Vector Machines (SVM) model. As such, the multilingual texts of unknown category can be classified by the trained classifiers. Subsequently, the resulting categorized texts were measured by means of a language-neutral clustering technique based on Self-Organizing Maps (SOM) method for evaluating semantic similarity among texts. The preliminary results show that our platform framework has the potential for cross-lingual text relatedness evaluation and plagiarism detection.