Topic Clustering With Cohesion Filtering for Event Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in efficiently identifying events from complex datasets with numerous data sources, often resulting in false positives, negatives, or inaccurate data, especially in electrical distribution systems, due to increased dimensionalities and noise, making it difficult to determine relevant events from extraneous datapoints.
Innovation Solution
A device and method utilizing a clustering module to group data into topic clusters and a cluster cohesion filter to remove datapoints based on their distance from the cluster center, employing threshold multipliers and temporal features to identify relevant events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple data sources are introduced to improve event detection accuracy, then measurement precision is improved, but device complexity increases due to increased dimensionalities
Solution Approach 1:
The patent segments the complex dataset into multiple topic clusters using clustering algorithms. Each cluster represents a specific event type or pattern, dividing the high-dimensional data space into manageable segments that can be processed independently, thereby reducing overall computational complexity while maintaining detection accuracy
Solution Approach 2:
The patent extracts and removes extraneous datapoints from each cluster based on distance metrics and thresholds. This extraction process eliminates noise and irrelevant data points that contribute to computational complexity without affecting the core event detection capability, thus resolving the contradiction between precision and complexity
2Reliability
If all datapoints are processed to ensure complete event detection, then reliability is improved, but productivity decreases due to increased processing time
Solution Approach 1:
The patent performs preliminary clustering of datapoints into topic clusters before detailed event detection. This preliminary organization groups related datapoints together, allowing subsequent processing to focus only on relevant clusters and their cohesive datapoints, thereby maintaining reliability while significantly improving processing efficiency
Solution Approach 2:
The patent applies different processing thresholds and criteria to different clusters based on their local characteristics. Each cluster can have customized distance thresholds and filtering parameters optimized for its specific event type, ensuring reliable detection for each local context while avoiding uniform over-processing across all data, thus improving overall productivity
Data Source
AI summary
A device for identifying events based on a dataset comprising datapoints. The device comprises a clustering module configured to cluster the dataset into topic clusters; and a cluster cohesion filter module configured to filter datapoints in each topic cluster based on each datapoint's datapoint-to-cluster distance.