Current data mining solutions are based on decoupled architecture. The data mining tools assume the data to be already selected, cleaned and transformed. Dr. S.K. Gupta has proposed a comprehensive architectural framework for the complete KDD process. The idea is to give a clean approach for knowledge discovery and integration of Database Management Systems with Knowledge Discovery Systems.
Such a framework also addresses the situations where mining requirements are focussed. The fallout of this architectural framework is "Intension Mining". Intension mining is logical extension of incremental mining with the guided mining paradigm in which the user states his mining requirements beforehand.
As per requirements, knowledge is accumulated as the database grows and used whenever actual mining is to be done. Intensional algorithms make maximum utilisation of the previous mined knowledge, while mining the incremented database. This leads to tight coupling of pre-mining, mining and post-mining operations and hence a cleaner approach towards knowledge discovery in databases.
In this thesis, we have implemented an intension mining system and mined association rules using that system.