MAQUI: Interweaving Queries and Pattern Mining for Recursive Event Sequence Exploration.

Area of Science:

• Computer Science
• Data Mining
• Information Systems

Background:

• Traditional event sequence analysis using only querying or mining algorithms is insufficient for complex data exploration.
• Analysts often recursively combine querying and mining for deeper insights, a process not well-supported by existing tools.

Purpose of the Study:

• To propose a framework for the interwoven process of querying and mining event sequences.
• To develop a user interface, MAQUI (Mining And Querying User Interface), that enables recursive event sequence exploration.

Main Methods:

• Developed a novel framework to describe the recursive interweaving of querying and mining.
• Created the MAQUI system to implement this framework, facilitating interactive event sequence analysis.
• Conducted two case studies with domain experts to evaluate MAQUI's efficacy.

Main Results:

• MAQUI enables analysts to recursively explore event sequences by interweaving querying and mining operations.
• Case study participants found the interwoven approach effective in articulating their analytical questions.
• The system facilitated the discovery of novel insights within the event sequence data.

Conclusions:

• The proposed framework and MAQUI system effectively support flexible and recursive event sequence exploration.
• Interweaving querying and mining is a valuable strategy for gaining deeper understanding and novel insights from complex event data.
• MAQUI demonstrates the practical utility of integrating these analytical techniques for domain experts.