A method and device for intelligent classification of litigation risks based on multi-model fusion

By using a multi-model fusion intelligent classification method, the system automatically analyzes the text of homeowner complaints against property management companies, identifies risk levels, solves the problem of low efficiency in manual judgment, achieves accurate risk warning and resource optimization, and reduces litigation risks.

CN121961252BActive Publication Date: 2026-07-17BEIJING DINGTAI ZHIYUAN TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
BEIJING DINGTAI ZHIYUAN TECH CO LTD
Filing Date
2026-03-30
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

When handling homeowner complaints, property management companies rely on human judgment, which is highly subjective, inconsistent in standards, and inefficient. This leads to high-risk complaints not being prioritized, missing the best opportunity for mediation, and increasing operating costs and reputational risks.

Method used

An intelligent risk classification method based on multi-model fusion is adopted. Key entities are identified through the BERT-BiLSTM-CRF model. Combined with the intensity of emotional conflict, legal correlation features and historical citation features, the risk of text complaints is automatically analyzed, and a historical event database and keyword matching are used for refined risk classification.

Benefits of technology

It enables the automatic and real-time extraction of multi-dimensional risk features from massive amounts of customer service text, transforming them into a proactive and precise risk warning mode, optimizing resource allocation, reducing the probability of conflicts escalating to litigation, and improving the level of intelligence in customer service and risk management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121961252B_ABST
    Figure CN121961252B_ABST
Patent Text Reader

Abstract

This invention relates to the field of data processing technology, and in particular to a method and apparatus for intelligent classification of litigation risk based on multi-model fusion. The method includes: preprocessing target text data; segmenting the preprocessed target text data into sentences and matching them with preset sentiment keywords to obtain sentiment conflict intensity features; extracting legal association features from the preprocessed target text data; establishing a historical event database and extracting historical citation features from the preprocessed target text data based on the historical event database, and determining the litigation risk level by combining the sentiment conflict intensity features, legal association features, and historical citation features; matching the preprocessed target text data with preset public issue keywords and preset third-party subject keywords respectively to determine the public issue matching degree and subject intertwining degree, and correcting the litigation risk level by combining the active association coefficient. This invention can effectively reduce the incidence of litigation.
Need to check novelty before this filing date? Find Prior Art