Abstract:

This work proposes a complete online learning environment which combines several aspects: categorization of students based on their abilities, learning styles and preferences, adaptability of the presented contents, recommendation of appropriate learning materials.

Except offering series of conventional services, abundant information available, message delivery, to effectively meet the needs, preferences, and different knowledge backgrounds the system offers the possibility to select the most adequate learning materials: text, images, video, audio, links to external websites.

Tag-based collaborative filtering is used for recommendation of learning materials to the student. In the process of finding appropriate materials to be recommended to the student, the system determines the degree of similarity of the tags most often used by the student with the words in the title, abstract and keywords of the papers.

Students are categorized based on the learning style (VART model) and all activities related to the interaction with the system, such as visited pages, teaching materials and external publications, tags and notes entered in the process of learning, ratings set to the teaching materials, etc.

An important aspect of this learning system is generating and recommending appropriate teaching materials, as well as appropriate tags and ratings. The list of recommended contents is generated by finding similar profiles and learning materials. Information retrieval algorithms are used to determine the similarity between student profiles and the similarity between used teaching materials.

The advantage of this learning system compared to traditional ones is the possibility to use knowledge about the domain and the teaching strategies to support individualized learning.

Keywords:

Tag-based collaborative filtering, learning style, categorization of students.