The Research And Application Of Intelligent Search Platform For Educational Resources

Taking "the research and application of the key technologies of network video online education and training based on cloud platform" as background, this paper studies the relevant technologies of the intelligent search of the educational resources based on text,including two aspects: the search of educational resources and the automatic classification of educational resources. In terms of search, this paper focuses on the built-in scoring rules of Solr search engine. To the problem that the titles and tags with higher resource characterization capabilities are set the same weight with contents, which leads to low utilization rate of information and low accuracy rate of the front row search results, give the resource titles and tags higher domain stimulating factor to update the normalization factor,and use the priority information of tag keywords to establish the new term frequency statistics model, which effectively ameliorates the resource comprehensive evaluation rules and improves the accuracy of the front row search results. In terms of classification, this paper focuses on the text-based automatic classification algorithm for research and improvement. Use the weighted sum of frequency of feature terms in resource titles,contents and tags to construct the transition model of feature terms, utilize position information of feature terms in tags to establish the importance coefficient, a new text quantitative model is proposed by combining the transition model and the importance coefficient, use quantitative results of resources to initialize the SVM classification algorithm for training multi-class classification model, the model is used to achieve the classification of resources. Ultimately integration of resources category structure will be achieved.First of all, the paper sketches the related theories, including the knowledge of search engine, text quantization, feature selection and text classification algorithm. Secondly, the overall frame of the intelligent search platform for educational resources is designed. The design and implementation of the Solr based vertical search engine system and SVM based resource automatic classification system are described in detail, and the Solr search engine technology, Chinese word segmentation technology and the training process of SVM are explored. Again, the modified Solr scoring rule based on the keyword priority and the improved automatic classification algorithm based on vector space model are expounded.Then, the search and classification functions of the system are tested and the results areanalyzed. Finally, the work of the paper is summarized.