Geo-Activity Zone Data Filtering for Entity Behavior Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data processing methods for activity-based intelligence struggle to accurately filter and correlate large volumes of data to predict entity behavior, particularly in geo-activity zones, due to missing geospatial tags and the need for efficient metadata correlation across diverse data sources.
Innovation Solution
A method and system that filter data into geo-activity zone cells by extracting metadata from streaming and static data, determining missing geospatial tags through profile data and region-based geo-topic models, and applying geo-tagging, entity resolution, and geo-fencing to correlate data within defined spatial, temporal, and contextual boundaries, using advanced analytics and machine learning techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data filtering and correlation is performed using traditional methods, then data processing can be completed, but accuracy in predicting entity behavior is insufficient due to missing geospatial tags
Solution Approach 1:
The patent introduces geo-topic models as an intermediary mechanism to infer missing geospatial tags. These models act as a bridge between available data and required geospatial information, using topic modeling techniques to predict location data that is otherwise missing from the dataset, thereby enabling accurate entity behavior prediction without complete geospatial tagging
Solution Approach 2:
The system performs self-service by automatically inferring and completing missing geospatial tags through geo-topic models and region-based inference. Rather than requiring manual geotagging or complete external data, the system autonomously fills information gaps using patterns learned from the data itself, enabling accurate filtering and correlation without external intervention
2Productivity
If advanced analytics and machine learning techniques are applied to filter and correlate large volumes of data, then predictive analytics capability is enhanced, but data processing complexity increases
Solution Approach 1:
The patent segments the complex data processing task into distinct modules: geo-topic model training, geospatial tag inference, data filtering, and entity behavior prediction. Each module handles a specific aspect of the analysis, allowing the system to process large volumes of data through manageable, specialized components rather than a monolithic complex system
Solution Approach 2:
The system performs preliminary actions by pre-training geo-topic models and establishing region-based inference rules before actual data filtering and prediction. This preprocessing step creates reusable knowledge structures that simplify subsequent data analysis, reducing the computational complexity required during real-time predictive analytics operations
Data Source
AI summary
An approach for filtering data is presented. A relationship between first and second entity-metadata elements specifying a person and a vehicle, respectively, and between the person and the vehicle is determined. Representations of the first and second entity-metadata elements are displayed within a regular polygon that includes locations indicated by a geospatial tag that includes location information about the person extracted from profile information describing the person and by other geospatial tags included in metadata obtained from data extracted from streaming data and data at rest. The metadata includes contextual information that specifies an activity included in a domain of knowledge associated with law enforcement. Based on hidden Markov and support vector machine models, a frequent pattern growth algorithm, and a Kohonen map, another activity of the person is predicted.