Business process residual activity sequence prediction method based on trajectory replay

• Main Authors: SUN XIAOXIAO, YU DONGJIN, YING YUKE, YANG SIQING
• Format: Patent
• Language: Chinese; English
• Published: 22.10.2021
• Subjects: CALCULATING; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention discloses a business process remaining activity sequence prediction method based on trajectory replay. The method comprises the following steps: firstly, simulating the execution condition of a process trajectory in a real environment by using a trajectory replay technology, and selecting a candidate prefix trajectory set most similar to the execution condition of a current process instance from a historical event log based on extracted execution context information; afterwards, quantifying the importance of each attribute in the process trajectory, that is, an attribute weight matrix is calculated; and finally, based on the attribute weight matrix, screening out a prefix track with the highest comprehensive attribute similarity from the candidate prefix track set, and taking a suffix activity sequence as a residual activity sequence of the current track. The method has the characteristics of high prediction similarity, wide applicability, strong robustness and the like, and can effectively solv