ΠΑΝΕΠΙΣΤΗΜΙΟ ΙΩΑΝΝΙΝΩΝ - ΤΜΗΜΑ ΜΗΧΑΝΙΚΩΝ Η/Υ & ΠΛΗΡΟΦΟΡΙΚΗΣ SITE MAP | ΠΑΝΕΠΙΣΤΗΜΙΟ ΙΩΑΝΝΙΝΩΝ - ΤΜΗΜΑ ΜΗΧΑΝΙΚΩΝ Η/Υ & ΠΛΗΡΟΦΟΡΙΚΗΣ ΕΛΛΗΝΙΚΑ | ΠΑΝΕΠΙΣΤΗΜΙΟ ΙΩΑΝΝΙΝΩΝ - ΤΜΗΜΑ ΜΗΧΑΝΙΚΩΝ Η/Υ & ΠΛΗΡΟΦΟΡΙΚΗΣ ENGLISH
Παρασκευή, 07/10/2016   -   Ομιλίες
Σεμινάριο Τμήματος με τίτλο:"User profiling and incentive provision in participatory sensing applications", Merkouris Karaliopoulos

Στο πλαίσιο της διοργάνωσης των σεμιναρίων του τμήματος, τα οποία θα συνεχιστούν και για το ακαδημαϊκό έτος 2016-2017, θα πραγματοποιηθεί την Παρασκευή 7/10/2016 και ώρα 12:00 στην αίθουσα Σεμιναρίων του Τμήματος Μηχανικών Η/Υ και Πληροφορικής, ομιλία με τίτλο "User profiling and incentive provision in participatory sensing applications". Ομιλητής θα είναι ο κ. Μερκούρης Καραλιόπουλος, Researcher, Department of Informatics, Athens University of Economics and Business.

ΠΕΡΙΛΗΨΗ

In our work we revisit the broader issue of how crowds of mobile end users end up contributing to certain mobile crowdsensing (MCS) tasks. In doing so, we depart in three different ways from practices and assumptions about the role of end users, which dominate the related literature. First, we approach the matching of users with MCS tasks as a task selection problem on the user side, i.e., it is users who select tasks instead of tasks being assigned to them, and cast it as a multi-attribute decision problem with many alternatives. Secondly, we take a much broader view of user heterogeneity, which caters not only for the diverse individual preferences or skills but also for the cognitive processes underlying the choices of end users. Finally, we explicitly acknowledge the bounded rationality of human decision making, as this results from cognitive but also extrinsic (e.g., time pressure) constraints.
We then propose two types of models for the way users select tasks to contribute to, both of them capturing the user heterogeneity and bounded rationality prescriptions. One of them, (multiclass) logistic regression models, are grounded in the area of statistics and machine learning, whereas the other, lexicographic cognitive heuristics, originate from long research threads in the fields of cognitive psychology and behavioral economics. We briefly present these models and describe how they can be trained with user-generated data from prior interactions with the MCS application. Finally, we discuss how these enhanced models of users behavior can direct a more efficient provision of incentives to them.

©2010 DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING - UNIVERSITY OF IOANNINA
created by data|SCIENCE

Valid XHTML 1.0 Strict Valid CSS!