Dataset Anonimyzation for Machine Learning: An ISP Case Study
Published in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2021
Recommended citation: Lelio Campanile, Fabio Forgione, Fiammetta Marulli, Gianfranco Palmiero, Carlo Sanghez, "Dataset Anonimyzation for Machine Learning: An ISP Case Study." Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2021. https://www.scopus.com/inward/record.uri?eid=2-s2.0-85125259327&doi=10.1007%2f978-3-030-86960-1_42&partnerID=40&md5=6d455ec0ee65b4ddced72d8d76f57eb6
Cited by: 0
Abstract: Internet Service Providers technical support needs personal data to predict potential anomalies. In this paper, we performed a comparative study of forecasting performance using raw data and anonymized data, in order to assess how much performance may vary, when plain personal data are replaced by anonymized personal data. © 2021, Springer Nature Switzerland AG.
Author Keywords: Attribute suppression; Character masking; Cryptography; Customer Premise Equipment; Data generalization; Hash; ISP; Logistic regression; Pseudo-anonymization; Random forest; WISP
Bibtex citation:
@ARTICLE{Campanile2021589,
author = "Campanile, Lelio and Forgione, Fabio and Marulli, Fiammetta and Palmiero, Gianfranco and Sanghez, Carlo",
title = "Dataset Anonimyzation for Machine Learning: An ISP Case Study",
year = "2021",
journal = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
volume = "12950 LNCS",
pages = "589 – 597",
doi = "10.1007/978-3-030-86960-1\_42",
type = "Conference paper"
}