Data mining is a dynamic and evolving field within Artificial Intelligence (AI), where ongoing research is dedicated to advancing techniques. Recently, it has found significant applications in Explainable AI and the broader domain of machine learning. Pattern discovery, Pattern discovery, a well-established topic in data mining, has diverse applications such as association rule mining, clustering, classification, and feature selection.
When combined with symbolic AI, pattern discovery offers a declarative framework capable of identifying different pattern types like frequent itemsets, high utility itemsets, gradual itemsets, among others.
In this presentation, I will explore diverse aspects of pattern mining, covering tasks from computing frequent and minimal rare patterns to high utility itemset mining problems using symbolic AI. I will demonstrate how a SAT-based framework can efficiently solve these problems, and I will also showcase how integrating a decomposition paradigm can help tackle scalability issues.
Finally, I will provide some perspectives to conclude the discussion.
