Research projects

Discovery group

Probabilistic Prolog (ProbLog)

Period: 06/2006-05/2008
Researchers: Hannu Toivonen, Sampo Yrjänäinen
Funding: Alexander von Humboldt Foundation, Department of Computer Science, HIIT

We first introduced and now develop ProbLog, a probabilistic extension of Prolog. A ProbLog program defines a distribution over logic programs by specifying for each clause the probability that it belongs to a randomly sampled program. The semantics of ProbLog is then defined by the success probability of a query. The key contributions so far are an effective solver for computing success probabilities, the formulation and a practical solution of the theory compression task for ProbLog, and an explanation-based learning setting using ProbLog.

Mining of biological databases (Biomine)    

Period: 3/2005-12/2007   
Researchers: Hannu Toivonen, Petteri Sevon, Lauri Eronen, Petteri
Hintsanen, Kimmo Kulovesi   
Funding: TEKES, Jurilab, Biocomputing Platforms, GeneOS (in collaboration with the Department of Medical Genetics at the University of Helsinki, Karolinska institutet in Sweden, VTT Biotechnology and CSC)

The project develops methods and tools for mining for information from public biological databases. Bio-scientists may use the developed search engine to enrich their data collections, discover previously unknown connections and analogies with public databases, and thus concentrate resources to the most promising leads for further study. The primary application area is analysis of candidate genes discovered in gene mapping. The project has studied how to represent biological data as graphs where the nodes represent various concepts (such as genes, proteins, tissues, phenotypes, cellular components) and the edges represent their known or predicted relations (such as the connection between gene and biological process reported in a gene database). The project has developed semantics and methods for defining and computing the strength of the connections between concepts, for isolating relevant sub-graphs, and for visualization purposes.

User context and privacy (Context)

Period: 1/2003-7/2007
Researchers: Kari Laasonen, Mika Raento   
Funding: Academy of Finland , Hecse, HIIT/BRU   

This project studied characterization and analysis of information about user context and its application in proactive computing. The research especially focused on issues pertaining to privacy, on algorithms for analysing positioning data, and on software engineering for context-aware applications. The research has been carried out in cooperation with a user interaction group in the HIIT Advanced Research Unit. The project produced the ContextPhone software that gathers, stores and transmits context information in ordinary S60 mobile phones. Furthermore, it can automatically annotate pictures taken with a camera phone with context data and transfer them to e.g. a web page. With the help of the software, the project has studied how the mediation of context data affects user communication, and developed methods for refining cell-based positioning data into a more graphical form. ContextPhone has been used as a research tool at Berkeley , MIT, and the University of Art and Design Helsinki, among other places. In October, Google bought Jaiku, which since 2006 had been developing mobile presence services mainly based on ContextPhone.