Indicators for advances in knowledge building - Application of content analysis tools to two sets of CSCL discourse data from two comparable classes

Authored by Professor Nancy Law & the Learning Community Project TeamThis is a seminar organized to report on the research outcomes of work conducted under the HKU Strategic Research Theme on Information Technology, within the area of Applying Data Mining Techniques to Novel Applications. This seminar presents the work in progress by a collaborative team comprising researchers from the Centre for Knowledge Science & Engineering Research, Beijing Normal University (CKSER) at Beijing Normal University and the Centre for Information Technology in Education (CITE) at the University of Hong Kong. Their research have centred on using content and interaction analysis to identify patterns of cognitive engagement and facilitation in computer supported collaborative learning (CSCL) contexts and the contribution of data-mining to building models of students’ developmental trajectory in knowledge building.CSCL has become a pedagogy of choice for many who believe that collaborative inquiry based learning is more effective in nurturing the kind of abilities needed for knowledge work in the 21st century. However, CSCL may not necessarily lead to effective learning. In the past decade or so, the volume of publications that provide data-driven insight for identifying patterns of cognitive engagement and facilitation based on CSCL discourse analysis has been far lower than the volume of CSCL discourse accumulated. A major part of the challenge in CSCL research is the difficulties in conducting systematic content analysis of the discourse data, which is crucial for understanding students’ learning progress. In this seminar, we will introduce some strategies for content and interaction analysis of CSCL discourse, the tools that the research team has built and the preliminary findings from applying the tools to the analysis of two sets of CSCL discourse as an illustration of how data mining can contribute to the assessment of knowledge building outcomes.VINCA stands for Visual INtelligent Content Analyzer, which is the content analysis tool jointly developed by CITE, HKU and CKSER, BNU. Currently, it includes the following functions: Data preparation to convert Knowledge Forum® discourse in html to database format, Keywords retrieval, Manual coding support, Linguistic database and tools for continuously improvable support to domain ontology, mapping of keywords, Learnable semi-automatic semantic coding, Content analysis augmented social network analysis, Novelty and similarity analysis, Influence degree analysis (semantic) of specified note(s) or person(s)published_or_final_versionCentre for Information Technology in Education, University of Hong Kon