Automatic Construction Of Courses Topology Based On Association Semantic

Automatic Construction of Courses Topology (ACCT) can provide an important reference model for some related systems such as the automatic arrangement for college teaching plan, the online sale and recommendation systems of electronic books, etc. The core task of ACCT is how to construct the topology relations among courses based on courses semantic. The task can be roughly divided into Chinese word segmentation, the key words extraction, the representation of course semantic, the representation of subject semantic and constructing the associated relations among courses. In this paper, some technologies have been used synthetically to automatically construct courses topology, which include Chinese word segmentation, association mining technology and Association Link Network. The main contents of this research are as follows:(1) A dynamic window split-based approach for extracting professional course key terms. Key words are the foundation in the representation of text semantic. The main task of this study is to extract key words from text to represent semantic of a course. The approach has explored the suitable length and suitable border of window to effectively guarantee the correctness of the extracted core words.(2) A semantic representation method for courses based on association semantic. The effective representation of courses semantic not only help student understand the course content on the whole, but also facilitate machine processing automatically. The representation of courses semantic need to further mine association relations among keywords to construct the course semantic based on the extracted core words.(3) A semantic representation method for a given subject based on courses semantic. The representation of subject semantic is the foundation for constructing the topology relations among courses. This study comprehensively utilize each course semantic to extract subject associated semantic. Thus, the rule base of subject associated semantic is formed and then subject semantic view is built.(4) A constructing method for courses topology based on the Association Link Network Model. Based on the representation of course semantic in the second part and the representation of subject semantic in the third part, these key words in the representation of course semantic are regarded as the semantic elements and these representations of courses semantic are regarded as the semantic nodes in this model. The rule base of subject associated semantic is used to construct the topology relations among semantic nodes (i.e., courses).The experiments show that the ACCT approach based on association semantic has rationality and validity. The innovation of this paper is that the technology of association semantic network is used to construct the topology relations among courses, which can rapidly and automatically construct the topology relations. The research achievements in this paper can be directly applied to some related fields such as the automatic arrangement for higher education plan, the recommendation system of book sales website and large-scale online learning website.