Affiliations: Instituto de Telecomunicações,
University of Beira Interior, Covilhã, Portugal | Zirak s.r.l., Italy
Note: [] Corresponding author: Joel J.P.C. Rodrigues, Instituto de
Telecomunicações, University of Beira Interior, Rua Marquês D'Ávila e Bolama, 6201-001
Covilhã, Portugal. E-mail: [email protected]
Abstract: In the last years, information flood is becoming a common reality,
and the general user, hit by thousands of possible interesting information, has
great difficulties identifying the best ones, that can guide him in his/her
daily choices, like concerts, restaurants, sport gatherings, or culture events.
The current growth of mobile smartphones and tablets with embedded GPS
receiver, Internet access, camera, and accelerometer offer new opportunities to
mobile ubiquitous multimedia applications that helps gathering the best
information out of an always growing list of possibly good ones. This paper
presents a mobile recommendation system for events, based on few weighted
context-awareness data-fusion algorithms to combine several multimedia sources.
A demonstrative deployment were utilized relevance like location data, user
habits and user sharing statistics, and data-fusion algorithms like the
classical CombSUM and CombMNZ, simple, and weighted. Still, the developed
methodology is generic, and can be extended to other relevance, both direct
(background noise volume) and indirect (local temperature extrapolated by GPS
coordinates in a Web service) and other data-fusion techniques. To experiment,
demonstrate, and evaluate the performance of different algorithms, the proposed
system was created and deployed into a working mobile application providing
real time awareness-based information of local events and news.
