Determining Process Model Using Time-Based Process Mining and Control-Flow Pattern

Abstract

Determining right model of business process from event log is the purpose of process discovery. However some problems i.e the inability to discover OR, noise IntroductionDetermining model of business process from event log is the main purpose of process discovery. Process discovery is a challenging task in process mining. It is a set of techniques which automatically construct a model of an organization`s current activities and its major activities variations. These techniques use event log of activities within an organization. The business process model is analyzed to show the complexity of issues in activities and how to solve them. These issues exist in any field, e.g. business Process discovery comes up with many algorithms, e.g. alpha, alpha+, alpha++ [6]. The alpha, alpha+, and alpha++ cannot deal with noise, incompleteness issues and OR conditional. Heuristic miner algorithms [7,9] come up to solve the noise problem. However, most of the algorithms are unable to find OR conditional model. The existing algorithm frequently discovers the OR conditional as AND parallel or XOR conditional. The thought of parallel model discovery will change the result of activities [8]. When "wait and see" behavior model synchronization is occured, it needs OR parallel to model the parallel split and join. The "wait and see" behavior model synchronization occured when the actor can choose only one activity, all activity, or more than one activity in parallel split and join. In this paper we proposed ideas to discover OR conditional within business process model.One of important things from process mining is the idea of completeness which is related to noise. Incompleteness leads to false parallel relations discovery, e.g the discovered parallel relation is XOR but the right parallel relation in business process is OR. The new representation of OR-split uses combination the existing XOR-split and AND-split to make the model easier to be analyzed [13]. In other hand temporal activity-based algorithm [8] and control-flow pattern can handle discovery of business process model with incompleteness and same frequency noise issues. Non-linear dependence in temporal activity-based algorithm is used to solve incompleteness problem since it can discover more relation than linear dependence. Controlflow pattern is used to solve same amounts of noise frequency issues because it discovers relation based on transaction function of activity, therefore it can choose non noise relation in business process model.

Graph-Based Algorithms for Discovering a Process Model Containing Invisible Tasks

“…This section explains how control-flow patterns [19] can capture elementary aspects of processes These patterns closely match the elementary control flow concepts used in this research. These patterns are a sequence and parallel patterns.…”

Discovering Process Model from Event Logs by Considering Overlapping Rules

“…and analyze business processes with decision points from event logs [5]. Nowadays, all of information systems have event logs which contain information, such as Case ID, activity, timestamp, resources [6].…”

AHP-TOPSIS for analyzing job performance with factor evaluation system and process mining

“…Process mining is a technique in terms of process management which supports the analysis of business processes based on event logs [9][10][11]. It is a set of techniques which automatically construct a model of an organization's current activities and its major activities variations [12]. To extract the information from the real executions constitutes one of the objectives of process mining [13,14].…”