Development of a smart learning analytics system using Bangla word recognition and an improved document driven DSS

A smart learning analytics system is proposed here. In the field of research, education learning analytics is manipulated as an optimistic way to render learning. In this proposed system, we put forward an idea regarding the design of learning analytics based on document-driven Decision Support System (DSS) and Bangla handwritten words recognition. The proposed learning analytics system aims at classifying the documents via syntactic analysis as well as categorizing them by semantic analysis of their contents. But the documents may be presented as printed format or hand written format. To choose the optimal feature vector for hand written words recognition (especially for Bangla basic characters) projection-based features are mostly efficient. Decision makers use DSS for decision making which is actually a computerized information system. We propose a route through the classified observations for providing the explanation which may improve the acceptance of the decision maker with a better correction rate using kNN algorithm. This system can certainly assist researchers for learning as well as decision making.