TY - GEN
T1 - Supporting Interpretability in Predictive Process Monitoring Using Process Maps
AU - Maita, Ana Rocío Cárdenas
AU - Fantinato, Marcelo
AU - Peres, Sarajane Marques
AU - Maggi, Fabrizio Maria
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.
PY - 2024
Y1 - 2024
N2 - Most predictive process monitoring approaches rely on machine learning techniques. These approaches predict, e.g., the outcome of a process case. As widely known, many machine learning techniques do not inherently provide insights in a useful format for business process experts to interpret the provided predictions and understand the logic used to derive such predictions. Recently, we proposed VisInter4PPM, a business-oriented approach to visually support interpretability in predictive process monitoring. In this paper, we apply VisInter4PPM to a loan request business process, whose behavior is represented in a real-world event log of a financial institution. This is a multiclass prediction problem where requests can be approved, declined, or cancelled. VisInter4PPM relies on the results of the SP-LIME interpreter to generate explanations about the influence of each business process activity on the case outcome. Thus, the SP-LIME results are visually projected onto a BPMN process model. The resulting process map shows which activities contribute to the predicted outcome and to what extent.
AB - Most predictive process monitoring approaches rely on machine learning techniques. These approaches predict, e.g., the outcome of a process case. As widely known, many machine learning techniques do not inherently provide insights in a useful format for business process experts to interpret the provided predictions and understand the logic used to derive such predictions. Recently, we proposed VisInter4PPM, a business-oriented approach to visually support interpretability in predictive process monitoring. In this paper, we apply VisInter4PPM to a loan request business process, whose behavior is represented in a real-world event log of a financial institution. This is a multiclass prediction problem where requests can be approved, declined, or cancelled. VisInter4PPM relies on the results of the SP-LIME interpreter to generate explanations about the influence of each business process activity on the case outcome. Thus, the SP-LIME results are visually projected onto a BPMN process model. The resulting process map shows which activities contribute to the predicted outcome and to what extent.
KW - Explainable machine learning
KW - Interpretable machine learning
KW - Predictive process monitoring
KW - Process mining
KW - XAI
UR - https://www.scopus.com/pages/publications/85200761006
U2 - 10.1007/978-3-031-64748-2_11
DO - 10.1007/978-3-031-64748-2_11
M3 - Conference contribution
AN - SCOPUS:85200761006
SN - 9783031647475
T3 - Lecture Notes in Business Information Processing
SP - 230
EP - 246
BT - Enterprise Information Systems - 25th International Conference, ICEIS 2023, Revised Selected Papers
A2 - Filipe, Joaquim
A2 - Filipe, Joaquim
A2 - Śmiałek, Michał
A2 - Brodsky, Alexander
A2 - Hammoudi, Slimane
PB - Springer Science and Business Media Deutschland GmbH
T2 - 25th International Conference on Enterprise Information Systems, ICEIS 2023
Y2 - 24 April 2023 through 26 April 2023
ER -