COPE: Interactive Exploration of Co-Occurrence Patterns in Spatial Time Series.

Area of Science:

• Data Science
• Geospatial Analysis
• Time Series Analysis

Background:

• Spatial time series data is prevalent in fields like economics and environmental science.
• Existing research focuses on identifying events within time series, but not temporal relationships between spatial events.
• Understanding these relationships is crucial for analyzing event formation, spreading, and spatial interdependencies.

Purpose of the Study:

• To propose a visual exploration framework, COPE (Co-Occurrence Pattern Exploration), for detecting temporal relationships between spatial events.
• To enable users to extract significant events from spatial time series data.
• To facilitate the discovery of repeated co-occurrence patterns between events at different locations.

Main Methods:

• Development of the COPE visual exploration framework.
• Extraction of user-defined events from spatial time series data.
• Detection and analysis of temporal co-occurrence patterns between events across multiple locations.

Main Results:

• COPE enables the identification of temporal relationships (same time, before, after) between spatially distributed events.
• The framework facilitates the discovery of recurring patterns in event occurrences across locations.
• Effectiveness and scalability were validated through case studies and expert reviews on real-world datasets.

Conclusions:

• COPE provides a novel approach for analyzing complex temporal relationships in spatial event data.
• The framework enhances understanding of event dynamics, spreading mechanisms, and spatial interdependencies.
• COPE offers a scalable solution for exploring co-occurrence patterns in large-scale spatial time series datasets.