An artificial intelligence decision processing system and method

By constructing a constraint modeling and admission determination module to generate an admission determination mask vector, the decision space is reconstructed, and only compliant decisions are transmitted to downstream algorithms. This solves the problem of non-compliant computation in AI algorithms, and achieves efficient utilization of computing resources and controllability of the decision system.

CN122310309APending Publication Date: 2026-06-30林昕梅
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610393205.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-27
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

Existing AI decision-making methods cannot restrict the computation path before the AI ​​algorithm is executed, which leads to non-compliant candidate decisions entering the computation state, resulting in resource waste and uncontrollable behavior of the decision system.

Method used

By constructing a constraint modeling module, an admission determination module, a decision space reconstruction module, and an algorithm interface module, an admission determination mask vector is generated and the index space of the candidate decision set is reconstructed. Only decisions that meet the constraints are transmitted to the downstream algorithm module.

Benefits of technology

It effectively reduces the computational load of AI algorithms, avoids non-compliant calculations, ensures the controllability of decision-making system behavior, and improves the accuracy of output results and the efficiency of resource utilization.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122310309A_ABST
    Figure CN122310309A_ABST
Patent Text Reader

Abstract

This invention provides an artificial intelligence decision processing system and method. The system includes a constraint modeling module, an admission determination module, a decision space reconstruction module, and an algorithm interface module. The constraint modeling module converts constraint information into a set of structured constraint objects. The admission determination module obtains a set of candidate decisions and extracts attribute features corresponding to the set of structured constraint objects for each candidate decision, generating a corresponding feature representation. Logical or vectorized logical operations are performed on the feature representations based on the set of structured constraint objects to generate an admission determination mask vector corresponding one-to-one with each candidate decision. The decision space reconstruction module reconstructs the index space corresponding to the candidate decision set based on the admission determination mask vector, generating a reconstructed index set that corresponds only to candidate decisions that meet the admission conditions. The algorithm interface module generates a candidate decision reference mapping table based on the reconstructed index set and provides the candidate decision reference mapping table to downstream algorithm modules. This invention also provides a method for implementing the above system. This invention can prevent downstream algorithms from encountering candidate decisions that do not meet the admission conditions, thereby avoiding waste of computational resources.
Need to check novelty before this filing date? Find Prior Art