Discovering Frequency-Weighted Utility Temporal Relationship Rules Using Frequent Pattern Tree

The data mining approach has introduced many challenges and opportunities to the database community. The primary goal of data mining is to discover hidden patterns in data that are valid, novel, potentially useful, and understandable. Traditional temporal association and relationship-rule mining cannot adequately address the challenges posed by real-time transactional databases, which often exhibit temporal features and time-validity periods. Currently, the extraction of high-utility recurrent patterns is a significant area of study in temporal data mining due to its ability to account for the frequency of item sets in time-based transactions and the varied utility rates of each item set. This manuscript proposes a novel procedure for discovering relationship rules on time-variant frequency-weighted utility-based data, using frequent pattern tree hierarchical structures, which provide substantial benefits in terms of time and memory usage. We propose an efficient tree-hierarchical structure for extracting high frequency-weighted utility-based time-variant item sets. Initially, we construct a novel frequency-weighted frequent pattern tree structure with crucial information about the utility of item sets. Subsequently, we mine the entire set of utility patterns. This approach compresses large time-variant datasets into smaller data structures while the utility FP-tree avoids redundant dataset scans, prevents the generation of a large number of candidate sets, and conserves search storage space. Our proposed approach demonstrates competence in extracting time-variant frequency-weighted high utility item sets.