Research And Application Of Information Retrieval Based On Ontology For Teaching Platform

With the continuous development of network technology, network teaching is attracting more and more attention and used more and more commonly. The network teaching has many advantages like flexible learning time and the acquisition of more knowledge in less time compared with traditional teaching method. The reason the teaching platform has those advantages is the module of teaching resource library. However, there are some questions in the construction of teaching resource library, such as less-than-ideal retrieve results and non-reusable teaching resource. These two issues are becoming the major reason for the encumbrances of further development.The reason of less-than-ideal retrieve results is there is no good way for retrieve. Many platforms provide retrieval way based on catalogue and keyword matching. The former is not suitable for abundant learning resource, and the latter cannot understand the connection of knowledge. To solve this problem, I propose the application of ontology-based information retrieve in network teaching platform. The issues of non-reusable teaching resource stem from the uncertain criteria of describing learning resource. To solve this problem, this article did some research about various the domestic and foreign learning resource standards and I adopted the LOM (Learning Object Metadata) as the criterion for describing learning resource.In conclusion, this article studies the following aspects:the knowledge about information retrieve and ontology, direct and indirect semantic expansion direct. Also this article improves the existing semantic similarity algorithm, and studies the Chinese word segmentation technologies and the learning object meta-data model. On the foundation research of the above content, I propose a semantic retrieve framework. Then I developed a prototype system based on the Siyuan network teaching platform based on the framework. In addition, I performed some experiments and give the following conclusion:the precision ratio of retrieve based on keyword matching is slightly higher than retrieve based on semantic ontology model and the recall ratio of retrieve based on semantic ontology model is higher than retrieve based on keyword matching.