Knowledge discovery in feedback system using pixel oriented technique

Abstract

To achieve excellence in education and produce quality students which can be absorbed by the organizations, feedback of students is collected to identify holes in the current teaching methodology and other services by institutions. A large amount of data is generated every semester and it is worthless to collect these data unless it is utilized for knowledge discovery. Visual data mining is most appropriate for representation when data size grows to Terabytes. Proposed solution is divided into various stages: Data collection, Data preprocessing, Data representation and Data mining. An online feedback system is developed in-house for data collection. Preprocessing stage retrieves relevant fields of the data base for representation. Preprocessed data is visually represented using pixel oriented technique which generates an image of entire data. In data mining state, the knowledge patterns are discovered from generated image. © 2010 Springer-Verlag Berlin Heidelberg.