Case affirmation method and system for performing causal inference based on correlation graph information

• Main Authors: GUO RUNDONG, LIU BAOCHEN, LI YUJUN, BEN XIANYE, HU WEIFENG, ZHAO SIWEN
• Format: Patent
• Language: Chinese; English
• Published: 03.05.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention provides a cause affirmation method and system for performing causal inference based on correlation graph information, and the method comprises the steps: obtaining the fact description of a case; constructing a causal graph according to the fact description of the case; performing causal discovery on the constructed causal graph by using a GFCI algorithm, and sampling to obtain a causal sub-graph; de-noising the sampled cause sub-graph, and adding the de-noised cause sub-graph into loss to obtain a cause affirmation result; wherein the step of constructing the causal graph comprises the steps of obtaining key words of a case by utilizing a KeyBERT algorithm, and clustering the key words of the case. The invention provides a method for performing cause and effect inference by using a cause and effect graph to perform cause and effect affirmation, so that unstructured information in case fact description is fully utilized, similarities and differences of different cases are better distinguished,