A Framework for Designing Knowledge Map using Text-mining method

Knowledge map is a representation tool to visualize knowledge sources and relationships among knowledge artifacts, which is considered as a core element of knowledge management system. To construct a knowledge map, we need to identify which category each newly registered knowledge artifact is mapped into, called structured knowledge. However, in view of fully utilizing obtainable knowledge, it is necessary to consider existing knowledge artifacts that are stored in previously established DB but not categorized. Thus an approach to incorporate unstructured knowledge as well as structured one into knowledge map turns out to be critical. With this end in view, we identified relationship between structured knowledge and unstructured knowledge using text-mining method. In our framework, we extract keywords set from each knowledge artifact by means of text-mining and compare it with each other to calculate similarity between knowledge artifacts. With this similarity, we can determine which knowledge artifacts are mapped into a certain location of knowledge map. To embody our framework, we also designed technology based knowledge map and developed web-based system in the case of a system integration company.