An Indexing Method of Continuous Spatiotemporal Queries for Stream Data Processing Rules of Detected Target Objects

Area of Science:

• Computer Science
• Data Science
• Artificial Intelligence

Background:

• Real-time spatiotemporal query processing is crucial for risk analysis in domains like defense.
• The Rete algorithm is a leading method for rule-based processing but suffers from search overhead with large datasets.
• Efficiently binding spatiotemporal rules to stream data is a key challenge.

Purpose of the Study:

• To address the performance overhead of the Rete algorithm in large-scale spatiotemporal query processing.
• To propose and evaluate a novel hashing indexing technique for Rete nodes.
• To enhance the efficiency of rule binding for continuous spatiotemporal data streams.

Main Methods:

• Implementation of a hashing indexing technique integrated with Rete nodes.
• Comparative performance evaluation against the standard Rete method (Drools).
• Measurement of processing time under varying numbers of rules, objects, and object distributions.

Main Results:

• The proposed hashing indexing method demonstrated a significant reduction in rule-searching overhead.
• Performance improvements of at least 18% were observed compared to the Drools implementation of the Rete algorithm.
• The method showed consistent performance gains across different data scales and distributions.

Conclusions:

• Hashing indexing is an effective strategy to optimize Rete-based spatiotemporal query processing.
• The proposed technique offers a practical solution for real-time risk analysis and decision-making with sensor data.
• This advancement is vital for improving the efficiency of combat vessel sensor data analysis.