Bridging the World's Knowledge to Individual Knowledge Using Latent Semantic Analysis and Web Ontologies to Complement Classical and New Knowledge Assessment Technologies

摘要

Using the theory of mental models to account for the process of human mental information processing and representation has become an important approach in the field of knowledge engineering. Most methods of knowledge assessment are based on graphical concept mapping techniques. Tools like MITOCAR complement the graphical approaches by externalizing knowledge on the basis of natural language. Although this approach has proven valid and reliable in a series of previous studies, the annotation of propositions is still a problem. In this technological study we found out that some of these methodological weaknesses concerning annotations in model based knowledge assessment and concept mapping can be compensated by using information from large text corpora, e.g. LSA or ontologies from the semantic web.