Research On Process Mining Based On Workflow

Process mining techniques aim at extracting knowledge from event logs that best describe the recorded behavior. One of the most important tasks in process mining is process model discovery. In discovering process models, an algorithm is designed to build a process model from a given event log. Genetic Process Mining based on process tree can handle different kinds of structures except duplicate tasks, while avoid the dead lock and live lock problem in traditional genetic process mining based on Petri net. One of the problems of GA was that the quality of the candidates, in the population, was low such that it reduced the chance of finding a perfect answer. Another problem was the fitness function based on weight will mislead the algorithm and it is hard to measure the value of weight.In this paper, a new model to discover process models has been proposed. A combination of Genetic Algorithm and Simulated Annealing has been used in this model and the process model is presented in the form of a process tree. Genetic Algorithms has previously been used in this context. Moreover SA has been used to improve the quality of candidates in the population.In addition, a new hierarchical fitness measure has been proposed to evaluate process models based on event logs. The problem of process discovery was addressed through a genetic algorithm with a new fitness function that takes into account both replay fitness,precision, generalization and simplicity. The new hierarchical fitness function based on the four quality dimensions of process model would guide the evolution of GA and produce a better process model. The searching of the process trees can be guided by the preferences of the user while ensuring correctness. In addition, the algorithm explicitly handles the noise of event logs in two phases: a pre-processing of the log and a post-processing of the mined model. Traces of the log that are infrequent would be removed and arcs of mined model that are less frequently than a certain threshold would be post-pruned. And in this way, the precision and simplicity of the mined model would be improved.Through the experimentation on balanced and unbalanced logs, it has been demonstrated that the proposed model outperformed in terms of rediscovering process models, compared to other approaches which are proposed in the literature, which was the result of better fitness evaluation and increased quality of individuals. It came to conclusion that using GA and SA in combination with each other can be effective in this context.