Publication

Abstract : Immense growth in communication has paved way for existence of information across the world in wide separated zones. There exists a need for a mechanism to render apt information to the needy from this enormous source of information. This mechanism is of high demand for educational purposes. Knowledge based cloud (Kloud) proposes a solution to combine together the information in different area, which is managed by several organizations. It then organizes them into different sections and hence providing a platform to furnish relevant information to people in search of it. The paper discusses about a method based on Naive Bayes algorithm to classify documents pushed into Kloud. A variation to this algorithm has been implemented by calculating term weight using converged weight method resulting in better accuracy and speed. A comparative study of proposed variance in classification algorithm against the actual algorithm was performed. Further we also implemented two subclassification algorithms namely hierarchical subclassification and subcategorization using document similarity method.