Abstract:High utility sequential pattern (HUSP) mining aims to mine actionable patterns with high utilities, widely applied in real-world learning scenarios such as market basket analysis, scenic route planning and click-stream analysis. The existing HUSP mining algorithms mainly attempt to improve computation efficiency while maintaining the algorithm stability in the setting of large-scale data. Although these methods have made some progress, they ignore the relationship between additional items and underlying sequences, which directly leads to the generation of redundant sequential patterns sharing the same underlying sequence. Hence, the mined patterns' actionability is limited, which significantly compromises the performance of patterns in real-world applications. To address this problem, we present a new method named Combined Utility-Association Sequential Pattern Mining (CUASPM) by incorporating item/sequence relations, which can effectively remove redundant patterns and extract high discriminative and strongly associated sequential pattern combinations with high utilities. Specifically, we introduce the concept of actionable combined mining into HUSP mining for the first time and develop a novel tree structure to select discriminative high utility sequential patterns (HUSPs) for downstream tasks. Furthermore, two efficient strategies (i.e., global and local strategies) are presented to facilitate mining HUSPs while guaranteeing utility growth and high levels of association. Last, two parameters are introduced to evaluate the interestingness of patterns to choose the most useful actionable combined HUSPs (ACHUSPs). Extensive experimental results demonstrate that the proposed CUASPM outperforms the baselines in terms of execution time, memory usage, mining high discriminative and strongly associated HUSPs.

An Efficient Parallel High Utility Sequential Pattern Mining Algorithm

HUSP-SP: Faster Utility Mining on Sequence Data

An Efficient Algorithm for Extracting High-Utility Hierarchical Sequential Patterns

A New Algorithm of Mining High Utility Sequential Pattern in Streaming Data

Fast Utility Mining on Sequence Data.

Fast Utility Mining on Complex Sequences

Mining Actionable Combined High Utility Incremental and Associated Sequential Patterns.

A Fast Algorithm for Hiding High Utility Sequential Patterns

A Survey of Parallel Sequential Pattern Mining

Mining Top-k High Average-Utility Sequential Patterns for Resource Transformation

High-Utility Sequential Pattern Mining with Multiple Minimum Utility Thresholds.

Mining Interesting Sequential Patterns using a Novel Balanced Utility Measure

Towards utility-driven contiguous sequential patterns in uncertain multi-sequences

TKUS: Mining Top-K High-Utility Sequential Patterns

Incremental Targeted Mining in Sequences

ProUM: High Utility Sequential Pattern Mining

HANP-Miner: High average utility nonoverlapping sequential pattern mining

Utility Mining Across Multi-Sequences with Individualized Thresholds

Mining Statistically Significant Patterns with High Utility

HEPM: High-efficiency Pattern Mining