Advances in Sequential Pattern Mining and Temporal Knowledge Graphs

The field of sequential pattern mining and temporal knowledge graphs is rapidly evolving, with a focus on improving efficiency, scalability, and interpretability. Recent developments have introduced novel algorithms and frameworks that enhance the mining process, such as incorporating average utility and user-defined query targets, and leveraging temporal graphs to model evolving knowledge. These advancements have significant implications for applications in log analysis, question answering, and dialogue consistency. Noteworthy papers include:

  • A study that introduced TAUSQ-PG, an algorithm for targeted high average utility sequential pattern mining, which demonstrates improved runtime and memory efficiency.
  • Research on LUSPM, which proposes a compact data structure and novel algorithms to efficiently mine low-utility sequential patterns, achieving excellent scalability and outperforming existing baselines.
  • The development of MemoTime, a memory-augmented temporal knowledge graph framework that enhances large language model reasoning through structured grounding and recursive reasoning, achieving state-of-the-art results on multiple temporal QA benchmarks.

Sources

Targeted Sequential Pattern Mining with High Average Utility

Efficient Mining of Low-Utility Sequential Patterns

Regular Expression Indexing for Log Analysis. Extended Version

AssoMem: Scalable Memory QA with Multi-Signal Associative Retrieval

A faster algorithm for efficient longest common substring calculation for non-parametric entropy estimation in sequential data

D-SMART: Enhancing LLM Dialogue Consistency via Dynamic Structured Memory And Reasoning Tree

RAG Meets Temporal Graphs: Time-Sensitive Modeling and Retrieval for Evolving Knowledge

MemoTime: Memory-Augmented Temporal Knowledge Graph Enhanced Large Language Model Reasoning

Built with on top of