Based On The Papers Of The Writing Style Characteristics Plagiarism Check Optimization Method

With the rapid development of the web technology, on-line database and information resource are increasing and more comprehensive, which provide a much easier approach to seek information such as data, survey and academic papers to help their academic work. However, at the same time, plagiarism is much commonly happening than ever before. Hence, to figure out an efficient technique or application to detect plagiarism is urgently necessary.We have already gained some achievements on plagiarism detection, including the design and implementation of word frequency based and relative unit density model based plagiarism detection algorithm. The word frequency based version is able to efficiently detect the kind of plagiarism using exact copy, and the relative unit density model is dealing with partly copy as complement. However, both of the work is unable to figure out the plagiarism which has changed the original text. Based on this, we now introduce a new target object, characteristic of one's writing style, to optimize the existing plagiarism detection application with scope of work including,1. Based on related studies domestic and abroad, this article compares the common techniques of writing style analysis and semantic dictionary, to figure out an appropriate technique and adapt it according to the needs of plagiarism application.2. Introduce the preliminary work, during which we designed and implemented a word frequency based and a relative unit density model based plagiarism detection algorithm. The paper also analyses the limitation of the 2 version which is the base for improvement.3. With summary of the existing techniques and preliminary work, this article presents an optimized technique based on writing style characteristic analysis, constructs a basic semantic dictionary of writing style characteristic, and describes the system structure and function module of the application.4. Use case study to present the application environment and show the improvement of the optimization. The writing style characteristic analysis based plagiarism detection method proposed by this paper is a complement and optimization of the preliminary work. By detecting plagiarism which changes the original text, this paper introduces a brand new idea and a solid base for.the team and our continual work on a more efficient preliminary detection to prevent the bad phenomenon.