Event Risk Detection Using Fixed Token Positions by Initiator

Solution Overview

Problem

Existing risk event detection systems struggle to accurately distinguish between event initiators in human-computer interaction scenarios, leading to poor classification of false information and increased risk events due to inconsistent event description information from multiple initiators.

Innovation Solution

A natural language processing technique that converts event description information into a token sequence with fixed token positions for each sub-token sequence, allowing a pre-trained text classification model to identify contradictions and determine if an event initiator provides false information, thereby assessing the risk of the event.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If event description information from multiple event initiators is processed without structured token positioning, then the processing complexity is reduced, but the measurement precision of detecting false information deteriorates

Engineering Contradiction:

Improvedetection accuracy of false informationVSAvoidcomplexity of information processing structure

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments event description information from multiple event initiators into separate sub-token sequences, with each initiator's information assigned to a specific segment. This segmentation enables the system to track and analyze information from each initiator independently, improving detection accuracy while maintaining manageable processing complexity through structured organization.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by assigning different token positions to different event initiators, creating distinct processing zones for each initiator's information. This allows the text classification model to apply initiator-specific analysis to each local segment, enhancing the precision of false information detection for individual initiators while preserving the overall system structure.

Inventive Principle:

Principle #3Local quality

2Reliability

If event description information from multiple event initiators is processed without structured token positioning, then the processing speed is maintained, but the reliability of risk event classification deteriorates

Engineering Contradiction:

Improvereliability of risk event classificationVSAvoidcomplexity of information processing structure

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments event description information from multiple event initiators into separate sub-token sequences, with each initiator's information assigned to a specific segment. This segmentation enables the system to track and analyze information from each initiator independently, improving detection accuracy while maintaining manageable processing complexity through structured organization.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by assigning different token positions to different event initiators, creating distinct processing zones for each initiator's information. This allows the text classification model to apply initiator-specific analysis to each local segment, enhancing the precision of false information detection for individual initiators while preserving the overall system structure.

Inventive Principle:

Principle #3Local quality

3Loss of information

If event description information from multiple event initiators is processed without structured token positioning, then the ease of operation is maintained, but the loss of information regarding initiator attribution deteriorates

Engineering Contradiction:

Improveinformation loss of initiator attributionVSAvoidcomplexity of token sequence structure

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent segments event description information from multiple event initiators into separate sub-token sequences, with each initiator's information assigned to a specific segment. This segmentation enables the system to track and analyze information from each initiator independently, improving detection accuracy while maintaining manageable processing complexity through structured organization.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by assigning different token positions to different event initiators, creating distinct processing zones for each initiator's information. This allows the text classification model to apply initiator-specific analysis to each local segment, enhancing the precision of false information detection for individual initiators while preserving the overall system structure.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS12585686B2Event detection and classification method, apparatus, and device

Publication Date: 2026.03.24 ALIPAY (HANGZHOU) INFORMATION TECH CO LTD

AI summary

Implementations of the present specification disclose an event risk detection method, apparatus, and device. The method includes: obtaining event description information provided by a plurality of different event initiators when the plurality of different event initiators each initiate a target event to a same event target party in a same event service; then converting, into a token sequence, a plurality of character sequences of the event description information provided by the plurality of different event initiators, the token sequence including a plurality of sub-token sequences each corresponding to a character sequence of event description information provided by an event initiator; setting a set of a first number of token positions for each sub-token sequence of the plurality of sub-token sequences, and sequentially placing characters in each sub-token sequence of the plurality of sub-token sequences at a corresponding set of the first number of token positions based on an order of each sub-token sequence; and determining, based on a corresponding sub-token sequence placed at each set of the first number of token positions, token information of an event initiator corresponding to each set of the first number of token positions, and a text classification model, whether the target event is at risk.