Design And Implementation Of Chinese Journal Personalized Search Engine

Today, with the rapid development of the Internet, search engine people widely used is also a quick update. Traditional search engine is difficult to adapt to the personalized needs of people. Traditional search engine services is based on the retrieval, different users in search for the same term will return the same result, different periods of the same user to search the same term, also returns the same result, it does not take the user’s requirements change into account. At present, the personalized retrieval, although there are a lot of research, the application of personalized search in Chinese journal search still few and far between. On the premise of this research topic of this paper arises at the historic moment.This paper based on the current status of the development of search engine, personalized retrieval technology as the background, analyzes the development status of Chinese academic search engine, sums up the current demand for Chinese periodical personalized search engine. Add user interest information to the academic journals in the search, to realize personalized information retrieval based on user interest.In this paper, based on the research status of personalized retrieval, and the characteristics of the existing data retrieval platform, the related technology research scheme is determined. On the basis of the index and full-text retrieval, application based on the LDA text clustering technology, access papers interest model and user interest model, combined with the personalized requirements of users, I establish a personalized retrieval system based on users’ interests. System overall is divided into two parts:based on Lucene search engine and personalized search design. Based on Lucene search engine design uses Lucene full-text information retrieval tool kit, combined with the Java language to realize common conventional search engine design. Personalized search design, the first application based on the LDA text clustering cluster analysis was carried out on the paper to get paper probability model and the author interests model, the paper probability model is added to the index in the library. After the user interests model, based on establishing database of user interest, to sort personalized search results.Through the actual test, proves that this system has realized the Chinese journals based on Lucene search engine, users can make ordinary retrieval, online browsing and downloading journal literature. Build user interests model, and implements the personalized retrieval of Chinese periodicals according to user’s interest degree, returns the personalized search results. According to the system requirement and the search engine technology evaluation standard, design system test cases, analyze the test results, verify the system up to the established standard.