Web Teaching Resources Acquisition Hyperlinks Can Collect Evaluation Study

Along with the rapid development of modern Information technology, Internet has become a very important tool int getting and utilizing knowledge. There are many resources on the web, yet those resources are widely distributed and very badly organized. Facing so much information, it’s a critical question for people to get interested resource rapidly and efficiently. Automatic acquisition of topic resources is realized by collecting hyperlinks on the net. And one of key questions about getting topic resources is to judge that whether a link should be collected.This paper designed a topic-link judging system based on Bayesian Network, the main function of which is to judge whether to collect a hyperlink during acquiring topic-instructional resources on the web, and to direct the web crawler to crawl and collect links.Firstly, this paper intruduced the concept and significance of acquiring topic instructional resources, compared the relevant theories and technologies, researched the distribution state of topic resources, and analized the factors that affect the establishment of hyperlinks.Then, this paper intruduced the basic concept of Bayesian Network, including the application area and other relevant theories and technologies, and offered a plan to judge hyperlinks based on Bayesian Network. Step1, abstract variables of Bayesian Network from the factors that can affect links’establishment. Step2, through structual learning and parameter learning, estiblish a Bayesian Network, Step3, use the network to judge and predict whether a link should be collected.After that, focused on the judgement of hyperlinks during collecting topic resources on the web, this paper selected and improved the methods and technologies which were used in estiblishing Bayesian Network. By combing and taking advantages of priori knowledge and experimental data, the caulation works could be reduced and the efficiency improved.Finally, this paper worked out a hyperlink-judging system based on Bayesian Network. More than5000hyperlinks were downloaded on mutiple websites to test the effectiveness of the system. Results of the experiments showed that Bayesian Network could be used in judging hyperlinks and achived a good effect.