차원축소 방법의 적용을 통한 자취군집화 개선

양한나, 송민석 · 대한산업공학회 춘계공동학술대회 논문집 · 2012

Process mining aims at deriving useful information from events logs. Recently, many organizations such as high-tech companies, hospitals, municipalities, etc. apply process mining techniques to analyze and improve business processes. Since, event logs from such organizations are usually large and the process mining results tend to be complex, the trace clustering method has been developed. It divides logs into several clusters by classifying cases according to their features. In this paper, we apply dimensionality reduction techniques (preprocessing methods) to improve the trace clustering technique. We have developed a prototype system and performed experiments to validate our approach. The result shows that dimensionality reduction techniques perform better in terms of processing time and average fitness values of clusters.

Cited by 0 works.

Authors