Likelihood-based multiple imputation by event chain methodology for repair of imperfect event logs with missing data
- 문서 유형
- 연구출판물 · 학술대회 발표논문
- 등재
- SCIEScopusSSCI
- 저널 순위
- Scopus Q1
- 저널
- Internet Interventions-The Application of Information Technology in Mental and Behavioural Healthpp.9 · 2019
- 피인용
- Scopus 24
- FWCI (Scopus)
- 2.23
- ISSN
- 2214-7829
저자
초록
© 2019 IEEE.The event log recorded through an information system may be missing for various reasons, which fact may result in an imperfect event log. Performing analyses using such an imperfect event log can seriously affect the quality of the obtained results. Therefore, analyses should be performed only after processing of the missing part in the imperfect event log. In the fields of data mining and statistical analysis, various methodologies have been developed to handle data with missing values, but there are not many studies dealing with incomplete event logs that have missing data in the field of process mining. In this paper, we propose a likelihood-based Multiple Imputation by Event Chain (MIEC) method for dealing with imperfect event logs with missing data. An experiment was performed using sample event logs, and a case study was conducted using a real steel manufacturing event log to verify our method. We expect the proposed method to repair the imperfect event log to a high level and to obtain analysis result with high quality even if there are many missing data.
키워드
Event ChainEvent Log qualityMultiple Imputation by Event ChainRepairing Event LogSingle Imputation by Event Relationship
식별자
- Scopus EID
- 2-s2.0-85071198348



로그인
연구물 목록으로