Novel Genetic Algorithm for Association Rule Mining with Multi-Objective Extraction for Bakery Database

Nowadays, the advancement in the technology has led to the enormous growth of data generated in the digital form. This leads to the situation where to extract interesting and useful knowledge from this vast amount of data becomes an attractive and challenging task. To help the situation, Data Mining techniques can be used which extract the relevant information from a large amount of data by using predictive and descriptive models. Discovering Association Rules is one of the Data Mining Techniques that is widely used today for the purpose of, say, guessing the frequent buying patterns. The most popular algorithms used for this purpose are Apriori and FP-Growth algorithms, other methods simply inherit the properties of any of the two. These techniques for Association Rule Mining generated a large number of rules, leaving the database analyst to go through all and find the interesting ones. These Algorithms alone are not able to extract interesting Association Rules efficiently. So, to improve on performance , this paper proposes a new approach towards Association Rule Mining that makes use of Genetic algorithm and evaluates the generated rules based on Multi-Objective Evaluation over the bakery database. It will find out which products are frequently brought together in bakery, and will show how the proposed system will overcome the drawbacks of traditional Apriori algorithm.