Towards Automatic Support for Augmenting Concept Maps with Documents

Abstract. Concept mapping has been used extensively in educational settings as a learning and teaching tool and, more recently, as a knowledge elicitation tool to capture expert knowledge for preservation purposes. In these contexts, electronic concept maps annotated with supplementary multi-media resources can provide a rich source of information. Unfortunately, it can be challenging for users to identify the right resources to attach, especially when dealing with large resource collections. This paper presents ongoing research on methods for easing the annotation task by automatically searching a document library for relevant documents and suggesting them as potential associations for concepts in a map. The paper begins by discussing how concept map structure can be exploited to automatically generate queries to a database of indexed documents, to search for documents to link to the concepts in a map. It then presents methods for indexing documents to improve the search results. The methods for generating queries and indexing documents have been evaluated using two pre-existing expert knowledge models, with encouraging results.