Based On The Semantic Features Of Paper Plagiarism Check Optimization Method Research

The phenomenon of plagiarism is very common and all kinds of research institutions have been focused on how to resolve the problem by effective means.The project team which I attended has done a lot of research and analysis on paper plagiarism detection since2005,and has completed a few versions of plagiarism detection systems which respectively are based on frequency statistics of key words, based on the relative unit density and based on the writing style characteristics. The former two algorithms have a high capability of plagiarism detection and location when the two papers have a high similarity on content, and these two methods have been confirmed that both have a high recall rate. The later version of plagiarism detection considered writing style characteristics, and optimized the original methods. Such detection method also made a satisfied progress on detection accuracy and system effectiveness. However, all the above detection methods do not take the emotional semantic into consideration which limit the detection accuracy when two paragraphs have different emotional tendency. Thus, we consider the emotional semantic characteristics to optimize the existing plagiarism detection method.This paper has the following five main points:1. Research and compare the emotional semantic characteristics calculation methods of both foreign and domestic ones, and analyze the limitation of existing detection method, confirm the feasibility of merging the emotional semantic technique and the existing algorithm.2. Based on the earlier work, we combine the related technique at home and abroad, and raise a new plagiarism detection method which based on emotional semantic characteristics. We established the emotional semantic dictionary and a few evaluation resource, design and describe the new process and detection algorithm.3. Describe both the technical solution of prototype system which based on the cloud computing platform and the detailed implementation method.4. Through the experiments and the instance of LunWenBaoPlagarism Detection System, we elaborate the effect of optimization of new detection algorithm which based on the emotional semantic characteristics, and validate the effectiveness of new method.The paper focused on the paper plagiarism detection method which based on the emotional semantic characteristics, we optimized the original method and analyze the emotional tendency in paper. The algorithm considered new elements and did further improvement which make more strike on the paper plagiarism and academic corruption.