Research projects
Adaptive computing research group
Semantic Interpreter Widened Experience (Stepwise)
Period: 9/2006-8/2007
Researchers: Patrik Floréen, Petteri Nurmi, Fredrik Boström, Tianyan Liu, Tiina-Kaisa Oikarinen, Péter Boda (Nokia), Akos Vetek (Nokia)
Funding: Nokia Research Center
In this subcontracted project for Nokia Research Center , we aim at improving context-aware services with the use of statistical methods. We have developed an adaptive user interface “Capricorn” for mobile widgets. The web-based and platform-independent interface uses various adaptive web techniques to help the user manage her widgets. To name a few examples, we use collaborative filtering to recommend suitable widgets and we dim infrequently used widgets. The system is also able to provide context information to widgets, which can then adapt their contents on the basis of this. Examples of such services are location-aware weather forecasts and location-aware news feed.
Widgets Sharing (WiSh)
Period: 9/2007-12/2007
Researchers: Kristiina Karvonen, Sini Ruohomaa, Petteri Nurmi, Marja Hassinen, Olli Immonen (Nokia)
Steering group: Patrik Floréen, Guido Grassel (Nokia), Kristiina Karvonen, Lea Kutvonen, Martti Mäntylä
Funding: Nokia Research Center
The Widgets Sharing (WiSh) project did research on trust models suitable for reputation management of a widget-sharing system. The objective of the reputation system is to encourage installing and using new widgets by helping the users make trust decisions.
Personalised Ubiservices in Public Spaces (PUPS)
Period: 8/2007-7/2009
Researchers: Patrik Floréen, Petteri Nurmi, Christoffer Björkskog, Fredrik Boström, Joonas Kukkonen, Eemil Lagerspetz,
Tianyan Liu,
Taneli Vähäkangas, Petri Myllymäki, Tei Laine, Matti Jokipii, Petri Kontkanen, Jussi Lahtinen, Jukka Perkiö, Tomi Silander, Hannes Wettig, Esko Kurvinen, Tero Hasu, Peter Peltonen, Petri Saarikko, Heikki Ailisto (VTT), Immo Heino (VTT)
Funding: Tekes, yritykset
In the project we design, implement and field trial prototype ubicomp applications that combine context-awareness with personalisation in order to provide a better user experience in everyday tasks in public spaces. We undertake probabilistic modelling, make user experience research, develop services and prototypes, and design business models. Our target time to market for full-fledged applications is a few years ahead. The applications will run on the user's own mobile devices.
The two-year project started in August 2007. The research partners are HIIT (ARU and BRU) and VTT. The company partners are Bitlips, Ekahau, Elisa, Idean Enterprises, K-Plus, Nokia, Ramblas Digital, Tuulia International, UpCode, as well as the Finnish Federation of the Visually Impaired.
The project has two application areas: shopping services for customers in Itäkeskus K-Citymarket (Massive) and an aggregate application for mobile context-aware and personalised digital media (AndroMedia). We have shopping data to our disposal. We have made data analysis of the data and we build user models from it. Different available applications by project partners are integrated into our services to be developed.
Optimising Data-gathering in Resource-constrained Networks (Geru)
Period: 1/2007-12/2009
Researchers: Patrik Floréen, Jukka Suomela, Topi Musto, Petteri Kaski
Funding: Academy of Finland
The Geru project carries out basic research on data gathering in sensor networks. A typical example of the kind of problems that we study is scheduling the activities of the nodes in a network. We may wish to prolong the lifetime of a battery-powered network by switching off the nodes according to a schedule, for example, or to satisfy some data transmission demands in a wireless network in the shortest possible time. Such problems are closely related to traditional combinatorial problems, such as partitioning and colouring graphs. Typically, these problems are computationally hard. Therefore, we focus on approximation algorithms, with which we are able to efficiently come up with a solution that is not necessarily optimal but provably near-optimal. In 2007 we have studied, among others, local algorithms. These are distributed algorithms in which the operation of a network node depends solely on the input that was available at system startup in the immediate surroundings of the node. Through our work, we have gained new knowledge of the kinds of computational problems that can or cannot be solved by local algorithms.
Trust For All (Trust4All)
Period: 10/2005-12/2007
Researchers: Patrik Floréen, Michael Przybilski, Taneli Vähäkangas
Funding: Nokia Research Centre
The EUREKA/ITEA project Trust4All studied the concept of trust in software architectures. The focus was on context-aware systems, and especially the trust and safety features of such systems. The research was based on the middleware for embedded systems that was developed in the Robocop project and extended in the Space4U project. The project was a joint effort of several European universities and companies. The research project at HIIT/BRU was subcontracted by Nokia Research Center . In 2007 we continued working on approaches for secure access requirements and integrated them into the Trust4All middleware. We evaluated our models and concepts in the form of demonstrators. In the final phase of the project the concepts and technologies were transferred to Nokia.