Hi everyone, I just wanted to give a pointer to Dataconda (available here), a new software (free for teaching and research purposes) that analyzes relational data. It is very useful in the phase of data preparation, when you have to build a "mining table" from a database. Most of the research on data mining is on how to analyze a flat table, but very little has been done on how to build the table from the raw data, which in 99% of the cases is in a relational DB.

Say that you have a table "Customers" and a table "Purchases" such that each customer has 0:n purchases. Say that customers are divided into "good" and "bad" customers and that you want to classify them accordingly. Dataconda allows you to find, for example, that a customer's goodness is determined by the number of times that she bought products that were popular among young customers without you having to build this attribute or to even suspect that this could be a good predictor.

I have a conference paper on this topic and I would love to engage more researchers in this area. I'm also using Dataconda in my business analytics course and I can share material and slides.

I need to help me of
how to use Dataconda for discovering Rules from data as CFD Miner Algorithm
thanks in advance

Hi,

Sorry for the late reply. To use Dataconda to find rules with the CFD Miner Algorithm, just load your tables, the relationships among them, and automatically generate a flat table. Then, you can mine the flat table with any data mining algorithm.

Follow this tutorial: https://www.youtube.com/watch?v=V8dhnddgXEo

Sorry, here is the link to the tutorial

please I need email of you for contact with you