Event Detection in Text Streams via Dynamic Burst Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing systems for event detection in text streams struggle with identifying issues in large volumes of data, particularly when events have smaller signatures and occur over longer periods, and fail to differentiate between events happening within similar time frames, often relying on fixed monitoring periods and requiring predefined keywords.

Innovation Solution

The implementation of dynamic monitoring periods, constrained-graph creation techniques, and community burst detection to analyze text streams, allowing for the identification of events over varying time frames and differentiating between similar events, using algorithms like Kleinberg's Burst Detection to identify bursts and cliques in the data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If fixed monitoring periods are used for event detection, then the system can operate with simple time windows, but it cannot detect events with smaller signatures that occur over longer periods

Engineering Contradiction:

Improvedetection period adaptabilityVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system implements dynamic monitoring periods that adapt to different event characteristics. Instead of using fixed time windows, the system adjusts the monitoring period based on the detected event patterns and signatures, enabling detection of both rapid events and slow-developing issues with smaller signatures across longer timeframes.

Inventive Principle:

Principle #15Dynamics

2Productivity

If manual analysis methods are used for text streams, then the system can maintain simplicity, but it cannot handle large volumes of text data from online sources

Engineering Contradiction:

Improvetext analysis throughputVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system replaces manual mechanical analysis with automated computational processing. Text streams from multiple online sources are processed automatically using algorithmic event detection methods, enabling the system to handle large volumes of data at high throughput without human intervention.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If existing event detection systems are used, then they can detect large events, but they fail to differentiate between different events happening within similar time frames

Engineering Contradiction:

Improveevent differentiation accuracyVSAvoidevent detection difficulty

Core Design Contradiction:

Measurement precisionVSDifficulty of detecting and measuring

Solution Approach 1:

The system segments events based on their unique characteristics, signatures, and temporal patterns. By dividing the detection process into distinct analytical components that examine different event attributes, the system can differentiate between multiple events occurring in similar timeframes and accurately identify each event's specific nature.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11501058B2Event detection based on text streams

Publication Date: 2022.11.15 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

A text stream source is accessed that includes a plurality of text content items. Unique word groupings are determined for the plurality of text content items. A burst detection algorithm is executed to determine word groupings that are currently bursting and that started within a specified time period. Based on the word groupings, an issue is determined based on identifying a set of texts forming at least one clique.