- Group leader: professor Hannu Toivonen
- Group homepage: http://www.cs.helsinki.fi/research/discovery/
During 2008-2009, the Discovery team has been mostly working on graph mining, developing methods and tools for information search and link discovery in large, weighted networks. The main application area is biomedicine. During the period, the team continued its work on the Biomine search engine for biomedical links. A new international project, Bison, was started, with similar goals and ideas but with less focus on a particular application.
The Pattern and Link Discovery team has no subgroups.
Research
Both Biomine and Bison build on the idea of representing large, heterogeneous information collections as labeled and weighted graphs, and then mining those for useful information. Within Algodan, this work is mostly within theme learning from and mining structured and heterogeneous data.
Our work in theme sequence analysis, has concerned haplotype analysis in human genetics. In 2008-9 we have not been active in this field, but we do maintain and distribute our haplotyping software HaploRec (published in 2006), which has about 120 registered users.