Semi-Supervised Discovery of DNN-Based Outcome Predictors from Scarcely-Labeled Process Logs
dc.contributor.author | Folino, Francesco | |
dc.contributor.author | Folino, Gianluigi | |
dc.contributor.author | Guarascio, Massimo | |
dc.contributor.author | Pontieri, Luigi | |
dc.date.accessioned | 2023-02-15T05:05:57Z | |
dc.date.available | 2023-02-15T05:05:57Z | |
dc.date.issued | 2022 | |
dc.description.abstract | Predicting the final outcome of an ongoing process instance is a key problem in many real-life contexts. This problem has been addressed mainly by discovering a prediction model by using traditional machine learning methods and, more recently, deep learning methods, exploiting the supervision coming from outcome-class labels associated with historical log traces. However, a supervised learning strategy is unsuitable for important application scenarios where the outcome labels are known only for a small fraction of log traces. In order to address these challenging scenarios, a semi-supervised learning approach is proposed here, which leverages a multi-target DNN model supporting both outcome prediction and the additional auxiliary task of next-activity prediction. The latter task helps the DNN model avoid spurious trace embeddings and overfitting behaviors. In extensive experimentation, this approach is shown to outperform both fully-supervised and semi-supervised discovery methods using similar DNN architectures across different real-life datasets and label-scarce settings. | de |
dc.identifier.doi | 10.1007/s12599-022-00749-9 | |
dc.identifier.pissn | 1867-0202 | |
dc.identifier.uri | http://dx.doi.org/10.1007/s12599-022-00749-9 | |
dc.identifier.uri | https://dl.gi.de/handle/20.500.12116/40222 | |
dc.publisher | Springer | |
dc.relation.ispartof | Business & Information Systems Engineering: Vol. 64, No. 6 | |
dc.relation.ispartofseries | Business & Information Systems Engineering | |
dc.subject | Deep learning||Outcome prediction||Process mining||Semi-supervised learning | |
dc.title | Semi-Supervised Discovery of DNN-Based Outcome Predictors from Scarcely-Labeled Process Logs | de |
dc.type | Text/Journal Article | |
gi.citation.endPage | 749 | |
gi.citation.startPage | 729 |