A self-evolving process entry routing closed-loop method and system based on user feedback

By constructing a two-dimensional dynamic thesaurus and a user feedback mechanism, the problems of user description discrepancies and rule rigidity in the process entry routing method are solved, achieving highly accurate adaptive process routing and reducing manual maintenance costs.

CN122332500APending Publication Date: 2026-07-03四川电力设计咨询有限责任公司
View PDF 6 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
四川电力设计咨询有限责任公司
Filing Date
2026-06-02
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing process entry routing methods cannot handle the differences in users' colloquial descriptions, resulting in low matching rates, a lack of closed-loop feedback mechanisms that prevent self-evolution, and rigid matching rules that cannot adapt to dynamic working conditions.

Method used

A two-dimensional dynamic lexicon of process keywords and fuzzy synonyms is constructed. Combining word segmentation technology with a hit rate weighted fusion mechanism, the lexicon and weights are updated in real time based on user feedback. A pre-trained semantic representation model and working condition requirement data are introduced for adaptive adjustment.

Benefits of technology

It improves the accuracy of process routing, realizes the system's self-learning capability, adapts to complex contexts and dynamic working conditions, reduces manual maintenance costs, and ensures that the optimal routing strategy is output under any working condition.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122332500A_ABST
    Figure CN122332500A_ABST
Patent Text Reader

Abstract

This invention relates to the field of intelligent information processing technology, and more particularly to a self-evolving process entry routing closed-loop method and system based on user feedback. The method constructs a two-dimensional dynamic lexicon containing process keywords and fuzzy synonyms; it segments the user's natural language to extract core keywords; it calculates the first and second hit rates of the core keywords and the two-dimensional lexicon respectively, and then weights and fuses them based on the two-dimensional weight coefficients to obtain a comprehensive matching score; it returns recommendation results to the user based on the score and ranking threshold and obtains interactive feedback data; finally, it dynamically updates the underlying lexicon based on the feedback data, and adaptively adjusts the weight coefficients and ranking thresholds in conjunction with business operation requirements data. This invention breaks through the barriers of traditional single exact matching, and achieves a self-evolving routing mechanism without manual intervention through data closed-loop, significantly improving the routing accuracy in complex contexts and dynamic business conditions.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • CN111899829A

  • CN111899829B

  • CN112699663A

  • CN114238568A

  • CN118468881A