Odsek za tehnologije znanja - Department of Knowledge
Download
Report
Transcript Odsek za tehnologije znanja - Department of Knowledge
Location of JSI
Europe
Slovenia
Micro-location of JSI
Ljubljana
Department of Knowledge Technologies
Jožef Stefan Institute
Jožef Stefan Institute - Profile
• Jožef Stefan Institute is the leading Slovenian national
research organization in the area of natural sciences
and technology
• The Institute was founded in 1949.
• The staff of around 700 people is specialized in research
of:
– information and communication sciences &
electronics,
– physics,
– chemistry, biochemistry & nanotechnology
– nuclear technology.
Jožef Stefan Institute - Organization
Research Departments
Physics (6 deps.)
Chemistry and Biochemistry (5 deps.)
ICT & Electronics (7deps.)
Nuclear technology (1 dep.)
Department of Knowledge Technologies (21 members)
Department of Intelligent Systems (18 members)
Centres
Infrastructure (6)
Educational (4)
Center for Knowledge Transfer in IT (3 people)
Technology and Consultancy (13)
Department of Knowledge
Technologies (Profile)
• Head: Prof. Nada Lavrac
• Staff: 19 researchers + 2 support + 6 external
• Historical note:
– Founded as Artificial Intelligence Group in 1979
– Artificial Intelligence Laboratory in 1985
– Department of Intelligent Systems in 1995 (AI & NLP Lab)
– Splitting into two departments (KT & IS) in 2004
• Close collaboration in Slovenia with:
– Intelligent Systems Department at J. Stefan Institute
– Center for Knowledge Transfer in IT at J. Stefan Institute
(Head - Mitja Jermol, previously Dr. Tanja Urbančič)
– Artificial Intelligence Lab at the University of Ljubljana
– Companies Quintelligence and Temida
– …forming a group of approx. 60 people working in AI.
Department of Knowledge
Technologies (main research areas)
Department of Knowledge Technologies
(main research areas)
– Machine learning/Data mining:
•
•
•
•
–
–
–
–
–
probabilistic learning in noisy domains (8),
text and web mining (5),
inductive logic programming (2),
equation discovery (3)
Knowledge management (4)
Ecological modeling (3)
Decision support systems (2)
Human language technologies (3)
Logic & cognitive science (1)
Other collaboration relevant to ECOLEAD
• LAKOS Laboratory at the Faculty for Mechanical Engineering,
Univ. of Ljubljana
– CIM, cybernetic models, intelligent manufacturing, networked
organisations and distributed manufacturing
• Faculty of Economics, Univ. of Ljubljana
– networked organisations, business, management and
organisation models, knowledge management
• several Slovenian Clusters (eKnowledge, automotive, ICT,
high-tech)
Department of Knowledge
Technologies (Assets)
• Software:
– Text-Garden – set of (mainly) Text-Mining tools based
on portable C++ library
– LINUS, RSD – relational rule learners
– CN2-SD, APRIORI-SD – propositional rule learners
– Largarnge/LagraMge – Equation Discovery systems
– DEXi – system for multi-attribute decision support
• Language Corpora:
– MultexEast
–…
International (non-EU) Collaboration
• Carnegie Mellon University, US
– Link Analysis, Text-Mining, Market Analysis
• Microsoft Research, UK
– Text-Mining, Link Analysis
• Stanford University & NASA Ames, US
– Equation Discovery, Ecological Modeling
• Osaka University, Japan
– Equation Discovery
• Tokyo University, Japan
– Human language technologies
Involvement in FP5 EU Projects
• R&D projects
– SolEuNet – Data mining and Decision support for Virtual
Enterprises (JSI coordinator)
– METAL – Meta Learning approaches (FET)
– CINQ – Query Languages for Data Mining (FET)
– GMonit – Data mining & Decision support in Ecology
• Research Networks
–
–
–
–
KDNet – Knowledge Discovery Network
KMForum – Knowledge Management
Nemis – Text-Mining network
EUnite – Machine Learning and Knowledge Discovery Network
Active FP6 projects
• Integrated Projects (IP):
– ECOLEAD – Cooperative Networked Organizations
– SEKT - Semantically Enabled Knowledge Technologies
• Specific Targeted Research Projects (STREP):
– ALVIS – Next Generation Search Engines
– SIGMEA – Data mining in Ecological Modeling
• Networks of Excellence (NoE):
– PASCAL – Machine Learning for Interfacing techniques
• Specific Support Actions (SSA):
– WYS-CEC - Women and Youth in Science
What ECOLEAD expects from JSI?
• Main JSI contributions:
– …delivering state of the art knowledge and providing
research in the broad field of Knowledge technologies
(WP2-WP5)
– …coordinating workpackage WP7 (Training)
– …developing software solutions based on Knowledge
technologies for WP2 and WP3
– …collaboration in other workpackages
What JSI expects from ECOLEAD?
• Main JSI expectations:
– to be member of the leading CNO initiative and recognizable
research project
– to become competitive with non-EU (US) related research
– to build successful commercial and research software to support
CNO
– to integrate with other related research areas
– to have a real impact on the development of virtual business
JSI group at the first ECOLEAD meeting
in Ljubljana