Creation of a Topic Maps-Based Wiki with an Article Similarity Measurement

Abstract

We created a pilot case of a Topic Maps-based wiki site to exchange ideas about children’s behavior in elementary schools. The topic map of this site consisted of the article and subject topic types. The subject topic type consisted of topics classified into areas such as “behavior,” “competence,” “field,” and “school time.” Each article was registered as an instance of the article topic type and associated with relevant subject topics. To measure the similarity between two articles and to relate articles on the basis of the similarity, we used the Tanimoto Similarity. To improve similarity-based retrieval, it was suggested that more specific subject topics characterize the articles.