Logo des Repositoriums

Auflistung nach Autor:in "Pontieri, Luigi"

1 - 2 von 2
  • Zeitschriftenartikel
    Semi-Supervised Discovery of DNN-Based Outcome Predictors from Scarcely-Labeled Process Logs
    (Business & Information Systems Engineering: Vol. 64, No. 6, 2022) Folino, Francesco; Folino, Gianluigi; Guarascio, Massimo; Pontieri, Luigi
    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.
  • Konferenzbeitrag
    A technique for information system integration
    (Information Systems Technology and its Applications, international conference ISTA'2001, 2001) Greco, Sergio; Pontieri, Luigi; Zumpano, Ester
    Nowadays, a central topic in database science is the need of an integrated access to large amounts of data provided by various information sources whose contents are strictly related. Often information sources have been designed independently for autonomous applications, so they may present several kinds of heterogeneity. Particularly hard to manage is the semantic heterogeneity, which is due to schema and value inconsistencies. In this paper, we focus our attention mainly on the inconsistency which arises when conflicting instances related to the same concept and possibly coming from different sources are integrated. First, we introduce an operator, called Merge Operator, which allows us to combine data coming from different sources, preserving the information contained in each of them. Then, we present a variant of this operator, the Extended Merge Operator, which associates the integrated data with some information about the process by which they have been obtained. Finally, in order to manage conflicts among integrated data, we briefly present a technique for computing consistent answers over inconsistent databases.