A Novel Process of Parsing Event-Log Activities for Process Mining Based on Information Content

Process mining has piqued the interest of researchers and technology manufacturers. aims to extract information from event activities their interdependencies events recorded by some enterprise systems. An system’s transactions are labeled based on content, such as an activity that causes occurrence another, timestamp between events, resource which transaction originated. This paper describes a novel process parsing event-log content (IC). The attributes, especially names, used describe flow processes systems, is grouped hierarchically hypernyms hyponyms in subsume tree. least common (LCS) values these names calculated, corresponding relatedness them obtained. These create fuzzy causal matrix (FCM) for activities, algorithm designed mine structural semantic relationships among using enhanced gray wolf optimizer backpropagation algorithm. proposed approach resistant noisy incomplete logs can be reflect structure behavior logs.