Publication Details
On the Similarity of User-specific Features on Mobile Devices
Loub Radek, Ing.
mobile forensics, fingerprinting, user specific features, similarity
In this paper, we show what kind of user-specific data obtained from a mobile device is appropriate for similarity measurement. The data have to be hardware independent, transformed in a normalized form so that we can compare e.g., bookmarks from different web browsers, and have to hold high entropy so that they can be used to distinguish different users. Selection of our set of user-specific data is based on previous research, our own survey among mobile users who were asked what application, settings, or functions they usually use, and our own experiments. For each source of data a different comparison function that respects specific features of the data is employed. The result of comparison is expressed as a similarity function that sums partial results of comparison functions weighted by entropy of the data.
@inproceedings{BUT168454,
author="Petr {Matoušek} and Radek {Loub}",
title="On the Similarity of User-specific Features on Mobile Devices",
booktitle="Proceedings of SPI 2017",
year="2017",
pages="1--11",
publisher="University of Defence in Brno",
address="Brno",
url="https://www.fit.vut.cz/research/publication/11376/"
}