Text Mining Services for Trialogical Learning

The paper describes the architecture of Text Mining services for the Knowledge Practices Laboratory1 (KP-Lab). The Trialogical Learning and Activity Theory, as a new paradigm of e-Learning, stands behind the KP-Lab system. This approach requires active goal-oriented participation of learners, sharing of learning materials (knowledge artefacts), distributed and highly customizable learning environment, and knowledge-based exchange of information (practices, artefacts) among learners. The Semantic Web Knowledge Middleware was designed as a service layer for storage, maintenance, and interfacing of semantically annotated knowledge artefacts. Since artefacts and their annotations contain descriptions in natural language, the Text Mining services were devised to provide an intelligent access and manipulation with repository of artefacts. Design principles and implementation of supervised and non-supervised text mining methods for the extraction of conceptual maps and classification of knowledge artefacts are discussed in this paper.