Self-Service Registry Log Builder: A Case Study in National Trauma Registry of Iran Publisher Pubmed



Yari Eili M¹ ; Vafadar S² ; Rezaeenour J³ ; Sharifalhoseini M⁴ Show Affiliations
Source: Methods of Information in Medicine Published:2022

Abstract

Background  Although the process-mining algorithms have evolved in the past decade, the lack of attention to extracting event logs from raw data of databases in an automatic manner is evident. These logs are available in a process-oriented manner in the process-aware information systems. Still, there are areas where their extraction is a challenge to address (e.g., trauma registries). Objective  The registry data are recorded manually and follow an unstructured ad hoc pattern; prone to high noises and errors; consequently, registry logs are classified at a maturity level of one, and extracting process-centric information is not a trivial task therein. The experiences made during the event log building from the trauma registry are the subjects to be studied. Results  The result indicates that the three-phase self-service registry log builder tool can withstand the mentioned issues by filtering and enriching the raw data and making them ready for any level of process-mining analysis. This proposed tool is demonstrated through process discovery in the National Trauma Registry of Iran, and the encountered challenges and limitations are reported. Conclusion  This tool is an interactive visual event log builder for trauma registry data and is freely available for studies involving other registries. In conclusion, future research directions derived from this case study are suggested. © 2022 Georg Thieme Verlag. All rights reserved.