Research Of E-learning System Based On The Personalized Knowledge Search

Within the more intensified trend of the network technology into universitiesnowadays, the society pay more and more attention to this new teaching requirementwhich has become the core values of the informatization construction in variousuniversities. In this context, E-learning system emerges as the time requires. It not onlyprovides the rich learning resources, but also the flexible learning ways, regardless oftime and space. Unlike the traditional teaching methods, most of the E-learning systemin which lack of the effective helped behavior only provide a platform forhuman-computer interaction. And it does not provide the knowledge retrieval function.When the learners want to search and get information, they will need to go to thefull-text search engine which only provide non-discriminatory and "thousands of peopleside" results for learners. This can lead learners lose interesting in learning at last.According to this, personalized knowledge search has become the goal pursued by thedevelopers of the online learning system. Personalized knowledge search engine couldbuild user interested model on the basis of the traditional research mode,which takesthe user's educational background, interests, search purposes into account and optimizesthe index, then reorders the results submitted and realizes the personalized researchservice.The paper intends to design and implement a E-learning system based onpersonalized search engine. The major work is done as follows:(1)The primary task is to build an experimental environment. This work consists oftwo parts: Firstly, it needs to set up E-learning platform. Establish your own E-learningsystem according to the requirements after referring to the open source's structure of theE-learning system.The successful results show that the function of the establishedsystem meets the online learning requirements. Secondly, we need to establish a Nutchsearch engine. Configure environment variables and debug the code in eclipse in orderto give the initial testing data. Then deploy the file to the tomcat server. Completedthese tasks, we can get the inquiry entrance of the Nutch search engine.(2)We can optimize Nutch search engine from two aspects. Firstly, it's theoptimization of the Chinese word segmentation in Nutch, just word segmentation inChinese aiming at directing at the Nutch search engine. This paper throws out asuggestion that using the imdict-chinese-analyzer module into Nutch in regard to thequestion of insensitive to Chinese.At the same time, the paper expands the function ofmodule from three aspects,and the experimental results shows that the expanded wordsegmentation module has more precise in segmentation work.(3) The paper specifically analyze the way of constructing the domain ontology,and the process of building the user's model, and on which basis it sets up the userinterested model. Due to the large range of the teaching resources, constructing of theoriginal model will be more complex. The paper is mainly focused on computer science of constructing ontology. In addition, according to ontology and the collection of keywords of users' interests as well as weight calculation, it sets up the users' interest model.(4)The paper achieves an E-learning system based on personalized search. Througha series of behavior optimization such as the Chinese word segmentation optimization,sorting optimization and the optimization of user interested model based on the courseof ontology, etc. The code source of Nutch will be repacked with ant tools and be putinto eclipse for debug. Then the Nutch search engine is embedded into the originalE-learning system. At last, it will complete the construction of the E-learning systembased on personalized search.