No abstract

The shortcomings of the traditional password-based authentication mechanism are becoming increasingly apparent as we transition from "one user - one device" to a richer "multiple users - multiple devices" computing paradigm. The currently dominant research direction focuses on on-device biometrics, which require sensitive information, such as images of a user's face, to be constantly streamed from a single recording source, often the device on which a user is getting authenticated. Instead, in this work we explore the possibilities offered by heterogeneous devices that opportunistically collect non-sensitive data in smart environments. We construct an IoT testbed in which we gather data pertaining to a person's movement in space, interaction with certain physical objects, PC terminal usage, and keyboard typing, and construct machine learning models capturing the person's behaviour traits. We commence our examination with models constructed from data sensed during a previously-completed task run and with such models we achieve up to 68% user identification accuracy (c.f. 7% baseline) among up to 20 individuals. Taking into account the limits of behaviour persistence we then revise our approach to continuously refine the model with the most recently sampled sensor data. This method allows us to achieve 99.3% user verification accuracy and successfully prevent a session takeover attack within 12 seconds with less than 1% of false attack detection.

Gesla in biometrični pristopi, na primer prstni odtisi, prevladujejo v svetu overjanja uporabnikov. Zaradi pomislekov glede varnosti, zasebnosti in priročnosti, znanost išče ustrezne nadomestke, vključujoč metode na podlagi vedenjskih vzorcev posameznikov. Kljub temu, da ti pristopi odpravijo večino obstoječih pomanjkljivosti, zaenkrat še niso dovolj natančni za splošno rabo. Glavni razlog za to je dejstvo, da uporabnikovo vedenje ni nespremenljivo. En izmed potencialnih vzrokov sprememb v obnašanju je spremenljiva kognitivna obremenjenost posameznika pri izvajanju različnih nalog. V tem delu na podlagi podatkov devetih prostovoljcev predstavimo raziskavo vpliva kognitivne obremenjenosti na vedenjske vzorce z analizo EEG in senzorskih podatkov. S pomočjo algoritma naključnega gozda razlikujemo med štirimi različnimi stopnjami kognitivne obremenjenosti. Dosežemo 89,79% in 76% natančnost na podlagi EEG in senzorskih podatkov, s čimer prikažemo povezavo med podatkovnima zbirkama.

No abstract