Predictive Modeling of Inter-Transaction Association Rules - A Business Perspective

Traditional association rules are mostly mining intratransaction associations i.e., associations among items within the same transaction where the idea behind the transaction could be the items bought by the same customer. In our work, we utilize the framework of inter-transaction association rules, which associate events across a window of transaction. The new association relationship breaks the barriers of transaction and can be used for prediction. Typically, the rules generated are derived from the patterns in a particular dataset. A major issue that needs more attention is the soundness of the rules outside of the dataset from which they are generated. When new phenomenon happens, the change in set of rules generated from the new dataset becomes more significant. In this paper, we provide a model for understanding how the differences between different situations affect the changes of the rules based on the concept of groups what we call factions. Also we provide a technique called Coalescent dataset, to generate set of rules for a new situation. Various experimental results are reported by comparing with real life and synthetic datasets, and we show the effectiveness of our work in generating rules and in finding acceptable set of rules under varying conditions.