An Interactive Exploration Learning Method Based on Domain Knowledge Visualization

Yuan La,Liu Yuan

DOI: https://doi.org/10.1109/aiam57466.2022.00063

2022-01-01

Abstract:A friendly, efficient and appropriate learning environment is the basic condition for improving the efficiency of learning in an informal learning environment. To address the issue of how to provide an effective learning environment in a non-systematic learning environment. In this paper, we propose an interactive exploration learning method based on domain knowledge visualization, we also designed a learning system for exploratory search of health information, taking the medical and health field as an example. Based on the construction of a domain knowledge base, an interactive knowledge visualization method is designed to complete the learning task in an interactive exploratory search process. In this paper, we use the MeSH vocabulary collection to construct a health domain ontology, and use the knowledge representation and reasoning capabilities of the ontology to realize interactive semantic query and data retrieval by combining with knowledge visualization techniques. Construct an interactive learning environment by storing health care knowledge and resources obtained from the Web in an elastic search database. In the interactive learning process, the improved BM25 algorithm is used to continuously rank the articles displayed on the page according to user behavior, determine the semantic similarity between keywords based on the Word2vec model, and use the dual association of document keywords and mesh topic word list as the navigation option to guide the user's search. The proposed method was validated using resources obtained from WebMD, and the experimental data and user feedback results proved the practicality of the method, which enabled users with unclear query/retrieval purposes to obtain satisfactory query results in exploring queries and enriching the users' existing knowledge system.