Benefits of Intension Mining

Just like a good database design can satisfy the user queries efficiently most of the times, a well thought of intension mining system will be able to discover different kinds of knowledge efficiently most of the times. The data miner can identify the type of knowledge & attributes of interest in the planning phase, which makes the mining process more efficient. Following are major benefits of intension mining :

1.

Reducing Search space : Search space can be pruned either by using heuristic or by taking into consideration the user's interest. Most of the times, the user's interest is limited to a few attributes, classes or patterns. As the attributes of interest are specified in the planning phase itself, the volume of data to be scanned reduces by a huge factor.

2.

Database scan reduction : In Intension Mining, Knowledge Concentrates are used for mining purposes, thus avoiding the scan of the database. Hence, planning the mining requirements beforehand shifts the thrust of data mining from being I/O intensive to computation intensive. Now, mining phase reads reduced data in a specific format & this has a visible speeding effect on mining.

3.

Integration with databases : The data intensive nature of data mining makes it imperative to design & implement methods to make the process faster. The current SQL interface is far too slow to be of any practical use in mining operations. The proposed architecture motivates work towards development of integration of database systems & knowledge discovery systems. Another advantage that the mining algorithms can draw from this integration is the use of semantics that are stored in database. These semantics can be used in conjunction with the domain knowledge to prune search space & get more reliable knowledge from data.

4.

Clean Approach : The disconnectedness in the various steps of Knowledge Discovery has been the chief motivating factor for evolving of Intension Mining. Intension Mining is a complete platform for data selection, cleaning, transformation, along with the mining and presentation capabilities. The basic idea is to insulate the user from the intricacies of the knowledge discovery process & hence giving him a clean approach for discovery of knowledge.

 

Figure: Architecture of Intension Mining System