الفهرس 
Mining unlabeled event log
AbstractOnly 14 pages are availabe for public view
 |  | from | 97 |  |  |
| | | from | 97 | | |
Abstract
Most of Information systems produce event logs as an evidence of the tasks that have been executed. If the execution of these tasks is controlled by a process execution engine, the resulting events are automatically correlated to each other. By correlation, we mean that the events which are executed by the same process instance, they have the same case identi{uFB01}er. Process min- ing uses these logs to perform process discovery and conformance checking with the original process model. It assumes that these logs have correlated events. Realistically speaking, it is rare to have a managed execution of busi-ness processes. Information systems that execute tasks are semi-automated systems or unwell integrated systems and have no central orchestration sys- tem. Thus, the generated event logs are uncorrelated and called unlabeled event logs. Manual preprocessing steps are required to correlate the events so that the process mining can proceed. Which is a tedious and error-prone task due to a large number of events that must be processed. In this research, we address the problem of correlating the unlabeled event logs that can{u2019}t be used directly in any process mining techniques. This problem has received little attention in the community of business pro- cess management. The existed approaches focus on de{uFB01}ning the executed behavior with a limitation of assuming that the process is an acyclic busi- ness process only. In this thesis, we focus on correlating the logs that are generated from the cyclic business processes and creating the labeled logs