The Case Base

One of the project’s objectives was to set up a case base of successful cases on the internet. The shared knowledge allows all internet users to benefit from a new case. Submitting a new case of best practice is a safe advertisement for KDD specialists or service providers, since the relational data model is kept private. Only the conceptual and the case model are published. Currently, there are four cases available: two for customer churn prediction, one for setting up customer profiles to focus a marketing campaign, and one for the prediction of sales of some items of a drug store chain.

To support users in finding the most relevant cases, their inherent structure is exploited. An internet interface was set up that visualizes the conceptual metadata. It is possible to navigate through the case base and investigate single steps, i.e. which operators were used on which kind of concepts. The internet interface reads the data directly from the M4 tables in our central database, avoiding redundancies.

Additionally to the data explicitly represented in M4, a few more informal descriptions are added to each Case that should give an overview of what the Case is about, what kind of data was used and how its goals were achieved.

If a user considers a case useful, then its conceptual data can be downloaded from the server (each case has been exported into a zipped M4 file which can be found in the case base; this file can be imported into the user's local database using the MiningMart system). If problems arise, or further help is necessary, contact information is provided about the case designer or the company that provided the service.

The case base can be found here.