The Copyright Protection Of Library Digital Resources Based On The Digital Fingerprint

With the development of information technology, library collections tend to be digitized. Constantly resources digitize make it convenient to users, but it’s easy duplication and propagation characteristic, also cause the problems that plagiarism and splicing of digital resources becomes more and more serious. Copyright of digital resources is under great threat.Resolve the copyright disputes of digital resources not only rely on the law, but also require technical support. Plagiarism can be detected by identification of digital resources resemblance. Traditional detection of similarity focus more on string matching and can not accurately determine plagiarism or not when facing advanced plagiarism such as synonym substitution. And traditional detection has the problems that computational complexity, can not reflect content of resources and is not suitable for massive data processing. So traditional detection now is gradually replaced by similarity detection method based on digital fingerprints. Digital fingerprint detection reduces computational difficulty through date dimensionality reduction and lay the technical basis for rapid and accurate digital resources on massive plagiarism.By integrating a variety of factors including the word correlation, word meaning, TF-IDF features and location factor, the algorithm in this paper firstly extracts literature subject terms and establishes digital fingerprints based on the literature content after processing the subject terms with SHA-1 coding dimension reduction. Secondly, it uses the trilateral bit commitment based on elliptic curve to protect the security of digital fingerprints. Finally, it compares the content-based digital fingerprints to realize the similarity detection of the literature so as to provide a powerful guarantee for the research on the copyright protection of literature digital resources. In order to support big data processing, this paper realizes the generation algorithm of digital fingerprints and the detection algorithm of similarity detection with the MapReduce Programming framework to improve the efficiency of calculation. Besides, in this similarity detection of the resources, this paper realizes the precise tracking of the source of the plagiarizing literature by comparing local digital fingerprints.Experiments show the digital fingerprint has obvious distinguish ability and strong robustness generated using the algorithm of this paper, the distance between the fingerprints can effectively perceive the changes of literatures content. Compared with the traditional similarity matching method, the detection more accurate and the computation speed is improved, can meet the requirement of similarity detection in big data, effectively solve the copyright violations of digital resources, and provide stronger protection for copyrighter.