The architecture of the intension mining system[1] is given on the previous page. For details of the architecture, please see the appendix.
The front-end layer in the architecture of intension mining system provides the user with the interface to the Knowledge Discovery Process. It typically consists of Knowledge Discovery Query Processor(KDQP), Knowledge Discovery Schema Editor(KDSE), Knowledge Discovery Schema Compiler(KDSC), and presentation tools. All these components are coordinated by a Front-end Engine (FEE). This layer is responsible for accepting all the user requirements and displaying the messages from the system. Facilities exist to enter and manipulate the schema and the mining query using KDSE and process it using KDSC.
The main issues at this level were the design of a Knowledge Discovery Schema Language(KDSL) i.e. the language for communicating the knowledge discovery requirements to the intension mining system. The language should be rich enough to capture all the user requirements along with the domain knowledge and focusing information.
User interaction is primarily at the planning phase of Intension Mining . It allows framing of data mining requirements and communicating the same to the system. So, a knowledge discovery schema editor was required which should be user-friendly and easy to be handled by a person who is not conversant with the intricacies of data mining.
Then, a knowledge discovery schema compiler was needed to compile the knowledge discovery schema. The input to the compiler could be the knowledge discovery schema entered through the knowledge discovery schema editor or though a text editor(if the user is very well conversant with the KDSL). The compiler validates the details specified by the user during schema compilation.
After that, there was a need to integrate the intensional mining algorithms to give a complete Intension Mining System.
In this thesis, we addressed the problem of specifying the knowledge discovery schema through a comprehensive KDSL in Intension Mining. Then, a knowledge discovery schema editor(KDSE) to edit that language and a knowledge discovery schema compiler(KDSC) to compile that language was to be implemented. After that an intensional algorithm for mining Association Rules[3] was to be integrated to give a complete Intension Mining System.