A tutorial search engine based on Bayesian learning

We present the prototype of a tutorial search engine that applies Bayesian learning to generate a ranked list of documents relevant to a given user query. The initial knowledge base used for training was obtained from university students input via the data collecting web page. The search engine is built around an on-going Java tutorial system and encapsulated by a web-based interface. The preliminary tests show a successful search rate of 90% - 100% accuracy by counting whether or not all five top search results are relevant to a user request.