The Security Research Of Search Engine Website Snapshot System

First the article introduces development status of search engine, discuss the future development trend. After expatiating on the related technology of search engines, it put forward the research of search engine website snapshot. Now search engines provide faster and more convenient visits to users by saving snapshots on massive cache, at the same time, snapshots may conserve unsafe code and return to the users when the users search, then the computers unknowingly infect with the virus. Therefore, this article not only realizes snapshots, but also carries on a security research.To realize the security judgment of snapshots, the article tried to use the artificial intelligence advanced method—machine learning. By combining the widely used cased-based learning and rote learning, it establishes the learning interpreter to judge javascript scripting language and remove unsafe coed or suspicious code.Now there are more and more malicious code, with the attendant, the possibility of snapshot unsafe code presence greatly increases. To identify malicious code, the system considers some feature keywords of familiar program segment qua example knowledge of machine learning repository, separates javascript code from html code, compares and judges it with the keywords of repository.Machine learning of the system through the implementation unit compare javascript program segment with repository. If the code is the same with or similar to the feature keywords, remove the segment from the original code after extracting feature keywords from similar code through study unit and storing these new examples in the repository to enhance the ability of learning system self-study. While the code which does not match the repository, it is considered as safe code and is reserved. Then the safety judgment snapshot system integrates with the chemical special search engine, improves on the system, and provide more secure snapshot to users.