@inproceedings{4f30edc3d1e84e808a3c2200aecaf97e,
title = "MLDStore: DNNs as similitude models for sharing big data",
abstract = "The amount of data grows exponentially with time and the growth shows no signs of stopping. However, the data in itself is not useful until it can be processed, mined for information and queried. Thus, data sharing is a crucial component of modern computations. On the other hand, exposing the data might lead to serious privacy implications. In our past research we suggested the use of similitude models, as compact models of data representation instead of the data itself. In this paper we suggest the use of deep neural networks (DNN) as data models to answer different types of queries. In addition, we discuss ownership of the DNN models and how to retain the ownership of the model after sharing it.",
keywords = "Big data, Deep neural networks, Similitude model",
author = "Philip Derbeko and Shlomi Dolev and Ehud Gudes",
note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2019.; 3rd International Symposium on Cyber Security Cryptography and Machine Learning, CSCML 2019 ; Conference date: 27-06-2019 Through 28-06-2019",
year = "2019",
month = jan,
day = "1",
doi = "10.1007/978-3-030-20951-3\_7",
language = "English",
isbn = "9783030209506",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "93--96",
editor = "Shlomi Dolev and Danny Hendler and Sachin Lodha and Moti Yung",
booktitle = "Cyber Security Cryptography and Machine Learning - 3rd International Symposium, CSCML 2019, Proceedings",
address = "Germany",
}