Self-Learning Event Concept Store for Dynamic Text Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high proliferation of information from media sources makes it troublesome to properly identify and detect new event types in real-time, as conventional methods lack the ability to build and recognize emerging event models effectively.
Innovation Solution
A self-learning event concept store system that uses text analysis to extract entities and topic vectors from data streams, compares them against stored event models, and validates new knowledge by determining frequency of occurrence, allowing for the detection and validation of new event models and their association with disambiguated entities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional event detection methods are used, then existing event models can be recognized, but new emerging event types cannot be detected
Solution Approach 1:
The system enables automatic detection and building of new event models through self-service mechanisms. The event concept store automatically learns from incoming data streams, extracts entities and topic vectors, and generates new event models without requiring manual intervention or complex configuration, thereby improving adaptability while maintaining simple operation
Solution Approach 2:
The system performs preliminary actions by pre-processing data streams to extract entities and topic vectors before event detection. The event concept store maintains pre-built event models and templates that are ready for comparison, enabling rapid detection of both known and emerging event types without delaying the detection process
2Measurement precision
If manual event model building is performed, then event detection accuracy improves, but processing time increases
Solution Approach 1:
The system uses copying by maintaining event templates and templates structures that can be replicated and adapted. When new event types are detected, the system creates copies of existing event model structures and populates them with extracted entities and topic vectors, enabling rapid model generation with high accuracy without manual building time
Solution Approach 2:
The system applies parameter changes by dynamically adjusting event model parameters based on extracted data. The event concept store automatically modifies event model attributes, thresholds, and characteristics based on the frequency and patterns of extracted entities and topic vectors, improving detection accuracy while processing occurs in real-time
3Reliability
If all information from media sources is analyzed, then complete event detection is achieved, but information overload increases
Solution Approach 1:
The system applies extraction by pulling out only the essential entities and topic vectors from data streams. The event concept store extracts and stores only the critical components needed for event detection, filtering out redundant information and maintaining a concise representation that ensures complete event detection without information overload
Solution Approach 2:
The system uses segmentation by dividing the event detection process into distinct modules: data stream reception, entity extraction, topic vector generation, event model matching, and new model building. The event concept store is segmented into separate tables for known events, uncategorized events, and validation queues, enabling reliable and complete detection while simplifying information processing through modular architecture
Data Source
AI summary
A system and method for detecting events based on input data from a plurality of sources. The system may receive input from a plurality of sources containing information about possible events. A method for event detection involves pre-processing and normalizing a data input from a plurality of sources, extracting and disambiguating events and entities, associate event and entities, correlate events and entities associated from a data input to results from a different data source to determine if an event has occurred, and store the detected events in a data storage.