Paper Number

ECIS2026-2294

Paper Type

SP

Abstract

The quality of unstructured data in process mining is often inadequate and is further undermined by suboptimal data preparation practices and human decision-making under time and cognitive constraints. This preliminary study combines qualitative interviews (n=4) and a quantitative survey (n=27) to identify key data quality issues in the pre-processing step of process mining, such as missing events, incorrect timestamps, and ambiguous activity labels. Drawing on insights from behavioral economics, nudging strategies (e.g., default options, progress feedback, social norms) are proposed to complement technical solutions and improve data quality without restricting user autonomy. Results highlight the potential of structured guidance and peer validation over competitive incentives. The study contributes insights into how subtle interventions can address decision-making challenges in data preparation, offering practical recommendations for tool design and automation in process mining.

Share

COinS
 
Jun 14th, 12:00 AM

Designing For Better Data: Applying Digital Nudging To Improve Data Quality In Process Mining

The quality of unstructured data in process mining is often inadequate and is further undermined by suboptimal data preparation practices and human decision-making under time and cognitive constraints. This preliminary study combines qualitative interviews (n=4) and a quantitative survey (n=27) to identify key data quality issues in the pre-processing step of process mining, such as missing events, incorrect timestamps, and ambiguous activity labels. Drawing on insights from behavioral economics, nudging strategies (e.g., default options, progress feedback, social norms) are proposed to complement technical solutions and improve data quality without restricting user autonomy. Results highlight the potential of structured guidance and peer validation over competitive incentives. The study contributes insights into how subtle interventions can address decision-making challenges in data preparation, offering practical recommendations for tool design and automation in process mining.