|
Main
research directions:
They range from
purely basic research to advanced RTD projects.
System
modeling and design (4 Dr., 7 eng.)
RTD
in this area was started in 1991, as a continuation to telecommunication
research. New business system engineering and reengineering methods
have been developed (J.Barzdins,
A.Kalnins, K.Podnieks,
A.Zarins et al.). The most important achievement is one of the most
up-to-date in the world system modeling tool GRADE.
The tool supports all major system development phases prior to the
implementation phase, with special emphasis on object modeling and
business process modeling. GRADE
development was started in 1991 as a joint project with a/s DATI (Latvia)
and Infologistik GmbH (Germany). Currently the maintenance and development
of the tool is performed by the Institute together with Infologistik
GmbH. The tool is being used by a lot of companies in the world, including
Siemens Nixdorf, Boeing, Fiat etc. In addition, an advanced graphical
web site analysis tool LINK
VIEWER based on sophisticated graph layout algorithms has been
developed and sold to many users.
The most visible achievement in recent years has been the development
of Graphical Modeling Environment (A.Kalnins,
J.Barzdins, U.Sarkans, K.Podnieks,
A.Zarins, E.Celms, J.Viksna et
al.) based on metamodels, transformations of metamodels, and Editor
Engine. The starting point of this research was the 4-th Framework
ESPRIT project No. 23287 "Application Development for the Distributed
Enterprise (ADDE)", where the Institute was responsible for development
of universal graphical modeling facilities. Now this environment is
used for rapid building of domain specific graphical modeling tools.
Besides that, the Institute is involved in developing a UML based
modelling methodology and building of models for specific domains
(J.Barzdins, A.Kalnins
et al.). Our areas of special interest are business modeling, conceptual
modeling and ontologies.
Another direction is development of mathematical models for specification
and analysis of systems with time constraints (K. Cerans).
Real time systems (5 Dr., 13
eng.)
The
Institute has long-standing traditions in the area of telecommunications
software. At the end of the 80ies an advanced for that time design
tool based on SDL language was developed. During the 90ies research
in this direction was continued, and automatic test case generation
methods based on SDL language were developed (A.Kalnins
et al.).
In recent years several new areas are added to pure software development
and testing such as digital signal processing in multiprocessor environment,
numerical methods for digital signal processing applications and printed
circuit board design. It allows to design and implement products in
a very wide range (M.Alberts, G.Linis, J. Kalnins, J.Martinsons, A.Gaujens,
N.Veselis et al.)
Main products under development are:
- Software and hardware for Wireless Access Network (Wireless
Local Loop),
- Real time audio system over Ethernet,
- High-level control systems of video cameras and other devices,
- Ultra-linear power amplifier and other electronic devices.
Simulation
of discrete event systems (1 Dr., 1 eng.)
A
unique discrete event simulation system SITA has been designed (G.Jonins,
J.Sedols). SITA is used for statistical analysis of different real
life systems.
Graph
drawing and graph theory (3 Dr., 4 eng.)
Another
priority area for the Institute is graph theory and graph drawing.
Very efficient incremental layout algorithms for interactive design
of graph-like diagrams are developed (P.Kikusts,
K.Freivalds, P.Rucevskis, K.Boitmanis et al.). These algorithms have
been awarded the first and third prize in the Graph-Drawing Contest'95
in Passau (Germany) and the first and third prize in the Graph-Drawing
Contest'99 in Prague (the Czech Republic). These algorithms are used
in the system modeling tool GRADE.
Now this research is continued by contract with Tom Sawyer Software
Corporation (USA). Besides that the research in traditional graph
theory is continued. The main topic is topological graph theory and
combinatorical maps (D.Zeps).
Machine
vision (1 Dr., 2 eng.)
The
machine vision group (P.Kikusts, K.Freivalds et al.) is developing
algorithms and systems for automatic object recognition in images
or video sequences as well as other technical vision data sources
including laser or light emitting diode (LED) systems.
Some of the applications include image filtering (http://www.gradetools.com/karlisf/image/filter.html),
video deinterlacing, automatic optical inspection system for printed
circuit boards.
The main product is Rain Sensor for car windscreen automatic cleaning.
The innovative aspects of this research are protected by the following
patents:
- D.Deberne, L.M.LeBa, V.Nikulenkovs, P.Ruchevskis. Procede de commande
de nettoyage d'une surface de plaque par controle optique dinamique,
et equipement de mise en oeuvre. 1998, Nr 98 160 97, VSEF 429, France.
- K.Freivalds, L.M.LeBa, V.Nikulenkovs. Automatically controlled cleaning
equipment for a plate surface presenting varied conditions of dirt
contamination and the procedure for implementation. 2000, EU patent.
- V.Nikulenkovs, P.Ķikusts, K.Freivalds. A method and equipment for
automatic optical detection and classification of a type of contamination
on a vehicle window surface. AP/Ap-P561/01, Latvia, 2001.
Inductive
Synthesis and Computational Learning Theory (6 Dr.)
A
research direction where our results are internationally recognized.
In the 70ies important results concerning inductive synthesis theoretical
limits were obtained (J.Barzdins, R.Freivalds, K.Podnieks).
In the 80ies a new model of inductive synthesis (so called "dots expressions")
was developed and corresponding inductive inference rules formalized
(J.Barzdins, A.Brazma).
At the beginning of the 1990s a new efficient inductive synthesis
algorithm based on term rewriting systems was found (G.Barzdins).
In recent years a new approach to practical inductive synthesis based
on hypothesis space restriction by means of attribute grammars has
been developed (U.Sarkans, J.Barzdins).
Research on recursion-theoretical level was also continued (R.Freivalds,
K.Apsitis, J.Viksna). Another research
direction in recent years has been the application of inductive synthesis
to biocomputing and genome informatics (A.Brazma,
J.Viksna, K.Cerans). In cooperation
with Helsinki and Bergen universities and the European Institute of
Bioinformatics a new algorithm is constructed for discovery of statistically
important patterns in biosequences. Currently this research is continued
at the European Institute of Bioinformatics (A.Brazma,
U.Sarkans).
Quantum Computation (2 Dr., 7 eng.)
Research in Quantum
Computation was started in 1998. It was proved that quantum finite
automata can be of exponentially smaller size compared with equivalent
deterministic or even randomized finite automata (A.Ambainis and R.Freivalds).
M.Valdats proved
that the class of languages recognized by 1-way quantum finite automata
with a bounded error is not closed under union (and under intersection
as well).
R.Freivalds
and A.Winter proved that quantum finite transducers can compute relations
not computable by deterministic or randomized finite transducers.
Natural
Language Processing (3 Dr., 7 eng.)
The
Artificial Intelligence Laboratory (http://www.ailab.lv
) has been working on automated translation tools for the Latvian
language (A.Spektors,
I.Skadina (Greitane))
since 1993. A limited Machine Translation model LATRA for translation
of stock market texts has been developed. In 1997 the Laboratory joined
the Universal Networking Language (UNL) project (A.Spektors, I.Skadina,
E.Milconoka).
As a result, the experimental grammar for deconversion of UNL texts
into Latvian has been developed (http://ai2.mii.lu.lv/decocgi.htm
). Recently research in MT has been concentrated on further elaboration
of syntax and semantic constituents within the project "Automated
synthesis of language independent text representation" funded
by the Latvian Council of Science.
The Artificial Intelligence Laboratory has initiated the collection
of Latvian resources and continues the development of corpus of the
written language (K.Levane, A.Spektors)
and the spoken data (I.Auzina). A corpus of the first printed Latvian
texts has been launched in 2003 (www.ailab.lv/senie).
One of the tasks carried out at the Laboratory concerns the development
of electronic dictionaries and data banks. Major results are the terminology
databank containing 115 000 Latvian terms and their translation equivalents
(E.Milconoka)
and the electronic version of the K.Mühlenbach and J.Endzelin
"Lettisch-deutches Wörterbuch" with search engine (A.Spektors,
E.Milconoka,
N.Gruzitis). Several electronic dictionaries are available on the
Web. A number of computer-aided teaching aids have been developed
(I.Auzina et al.) for "Latvian Education Informatization System".
Significant
publications
1.Andris
Ambainis, Kalvis Rusins Freivalds, Carl H. Smith. Hierarchies of probabilistic
and team FIN-learning. - Theoretical Computer Science, 2001, v.261,
No.1, p.91-117.
2.A.Ambainis, R.Bonner, R.Freivalds, M.Golovkins, M.Karpinski. Quantum
finite multitape automata. - Lecture Notes in Computer Science, 1999,
v.1725, p.336-344.
3.A.Ambainis, R.Bonner, R.Freivalds, A.Kikusts. Probabilities to accept
languages by quantum finite automata. - Lecture Notes in Computer
Science, 1999, v.1627, p.174-183.
4.Andris Ambainis and Rusins Freivalds. 1-way quantum finite automata:
strengths, weaknesses and generalizations. - Proceedings of the 39th
Symposium on Foundations of Computer Science, Palo Alto, California,
November 1998, p.332-341.
5.Andris Ambainis, Arnolds Kikusts, and M. Valdats. On the Class of
Languages Recognizable by 1-Way Quantum Finite Automata.- Lecture
Notes in Computer Science, 2001, v.2010, p.75-86.
6.K.Apsitis, S.Arikawa, R.Freivalds, E.Hirowatari, C.H.Smith. On the
inductive inference of recursive real-valued functions. - Theoretical
Computer Science, 1999, v.219, No.1, p.3-17.
7.Kalvis Apsitis, Rusins; Freivalds, Raimonds Simanovskis, Juris Smotrovs.
Closedness properties in ex-identification. - Theoretical Computer
Science, 2001, v.268, No.2, p.367-393.
8.J.Barzdins, I.Etmane, A.Kalnins, and K.Podnieks. Towards integrated
computer aided systems and software engineering tool for information
systems design. - Advances in Databases and Information Systems, Springer
Verlag, 1996, p.3-11.
9.J.Barzdins, R.Freivalds, C.H.Smith. Learning with confidence. -
Lecture Notes in Computer Science, 1996, v.1046, p.207 218.
10.J.Barzdins, R.Freivalds, C.H.Smith. Learning formulae from elementary
facts. - Lecture Notes in Computer Science, 1997, v.1208, p.272-285.
11.J. Bārzdiņš, R. Freivalds, and Carl H. Smith. Towards axiomatic
basis of Inductive Inference. - Lecture Notes in Computer Science,
2001, v.2138, p.1-13.
12.J.Barzdins and A.Kalnins. Business system reengineering: tools
support. - Proceedings of the Fourth International Conference and
Exhibition "Informatics Technologies and Telecommunications in
the Baltic States", 1997, p.69-73.
13.J.Barzdins and U.Sarkans. Incorporating hypothetical knowledge
into the process of inductive inference. - Lecture Notes in Computer
Science,1996, v.1160, p.156 168.
14.J.Barzdins and Kalnins. How to comprehend large and complicated
systems. - Information Systems Development: Advances in Methodologies,
Kluwer Academic, 2002, p.215-226.
15.Aija Bçrziòa and Richard Bonner. Ambainis-Freivalds'
algorithm for measure-once automata. - Lecture Notes in Computer Science,
2001, v.2138, p.83-93.
16.Richard Bonner, Rusins Freivalds, and Maksim Kravtsev. Quantum
versus probabilistic one-way finite automata with counter. - Lecture
Notes in Computer Science, 2001, v.2234, p.181-190.
17.A.Brazma. Efficient learning of regular expressions from approximate
examples. - Computational Learning Theory and Natural Learning Systems,
MIT Press, 1996, v.4, p.351-366.
18.A.Brazma, K.Cerans. Noise-tolerant efficient inductive synthesis
of regular expressions from good examples. - New Generation Computing,
OHMSHA LTD and Springer Verlag, 1997, No.15, p.105-140.
19.A.Brazma, I.Jonassen, I.Eidhammer, D.Gilbert. Approaches to automatic
discovery of patterns in biosequence. - Journal of Computational Biology,
1998, v.5, No.2, p.277-303.
20.A.Brazma, I.Jonassen, J.Vilo, E.Ukkonen. Pattern discovery in biosequences.
- Lecture Notes in Computer Science, 1998, v.1433, p.255-270.
21.K.Cerans, J.Viksna. Deciding reachability for planar multi-polynomial
systems. - Lecture Notes in Computer Science,1996, v.1066, p.389-400.
22.Lance Fortnow, Rusins Freivalds, Wiliam I.Gasarch, Martin Kummer,
Stuart A.Kurtz, Carl H. Smith and Frank Stephan. On the relative sizes
of learnable sets. - Theoretical Computer Science, 1998, v.197, No.1-2,
p.139-156.
23.K.Freivalds. Curved edge routing. - Lecture Notes in Computer Science,
2001, v.2138, p.126-137.
24.K.Freivalds. A nondifferentiable optimization approach to ratio-cut
partitioning.- Lecture Notes in Computer Science, 2003, v.2647, p.134-147.
25.K.Freivalds, U.Dogrusoz, P.Íikusts. Disconnected graph layout
and the polyomino packing approach. - Lecture Notes in Computer Science,
2002, v.2265, p.378-391.
26.K.Freivalds, P.Íikusts. Optimum layout adjustment supporting
ordering constraints in graph-like diagram drawing. - Proceedings
of the Latvian Academy of Sciences, Section B,2001, v.55, No.1, p.43-51.
27.Rusins Freivalds. Quantum computers and computer science. - Proceedings
of the Third International Baltic Workshop on Databases and Information
Systems, 1998, v.1, p.8-14.
28.Rusins Freivalds.
How to Simulate Free Will in a Computational Device. - ACM Computing
Surveys, 1999, v.31, No.3, p.15.
29.Rusins Freivalds. Quantum computers and quantum automata. - Numbers,
Information and Complexity, 2000, p.541-553.
30.Rusins; Freivalds and Carl H.Smith. Category, measure, inductive
inference: a triality theorem and its applications. - Lecture Notes
in Computer Science, 2002, v.2286, p.386-399.
31.R.Freivalds and S.Jain. Kolmogorov numberings and minimal identification.
- Theoretical Computer Science, 1998, v.188, No.1-2, p.175-194.
32.Rusins Freivalds and Andreas Winter. Quantum finite state transducers.
- Lecture Notes in Computer Science, 2001, v.2234, p.233-242.
33.D.Gilbert, G.Torrance, J.Vîksna, N.Dhulia, D.Mamtora, D.Westhead,
I.Michalopoulos. A fast pattern-matching system on a protein topology
database. - Currents in Computational Molecular Biology, 2002, p.70-71.
34.Marats Golovkins. Quantum Pushdown Automata. - Lecture Notes in
Computer Science, 2000, v.1963, p. 335-345.
35.I.Greitane. Translation and multilingual resources for Latvian.
- Proceedings of the Third European Seminar "Translation Equivalence",
1998, p.79-86.
36.G.Jonins, J.Sedols. The overload - the causes, models, chances
to prevent. - Proceedings ITC Sponsored St.Petersburg Regional International
Teletraffic Seminar "Teletraffic Theory as a Base for QOS: Monitoring,
Evolution, Decisions", 1998, p.121-128.
37.G.Jonins, Dz.Tomsons. Simulation-based training using SITA. - Proceedings
of Eurosim'98 Simulation Congress, 1998, v.3, p.596-598.
38.A.Kalnins, J.Barzdins, E.Celms et al. The first step towards generic
modeling tool.- Proceedings of the Fifth International Baltic Conference
on Databases and Information Systems, Tallin, 2002, v.2, p.167-180.
39.A.Kalnins, J.Barzdins, K.Podnieks. MiniGRADE - a tool for conceptual
modeling by class diagrams. - Proceedings of the 18th International
Conference on Conceptual Modeling (ER'99), 1999, p.11-12.
40.A.Kalnins, K.Podnieks, A.Zarins, E.Celms, and J.Barzdins. Editor
definition language and its implementation. - Lecture Notes in Computer
Science, 2001, v.2244, p.74-81.
41.P.Kikusts. Mathematical principles of graph-like diagram drawing
tool. - Acta Societatis Mathematicae Latviensis, 2000, No 3, 47-48.
42.P.Kikusts, P.Rucevskis. Layout algorithms of graph-like diagrams
for GRADE windows graphical editors. - Lecture Notes in Computer Science,
1027 (1996), p.361-364.
43.Kristine Levane, Andrejs Spektors. Morphemic Analysis and Morphological
Tagging of Latvian Corpus. - Proceedings of the Second International
Conference on Language Resources and Evaluation.
Athens, Greece 31 May - 2 June 2000, 2000, v.2, p.1095-1098.
44.G.Linde. Reduction of UML class diagrams. - Proceedings of the
Fifth International Baltic Conference on Databases and Information
Systems, Tallinn, Estonia, 2002, v.2, p.39-48.
45.G.Linde. Reduction of UML class diagrams. - Databases and Information
Systems II, Kluwer Academic Publishers, 2002, p.199-208.
46.J.Rung, T.Schlitt, A.Brâzma, K.Freivalds, J.Vilo. Building
and analysing genome-wide gene disruption networks. - Bioinformatics,
2002, v.18, Suppl.2, p.202-210.
47.O.Sand, T.Vu, Tien Deng, D.Gilbert, J.Vîksna. Analysis of
Topological Representations of Transcriptional Regulatory Regions.
- Currents in Computational Molecular Biology, 2002, p.163-164.
48.U.Sarkans, J.Barzdins. Using attribute grammars for description
of inductive inference search space. - Lecture Notes in Computer Science,
1998, v.1501, p.400-408.
49.U.Sarkans, J.Barzdins, A.Kalnins and K.Podnieks. Towards a metamodel-based
universal graphical editor. - Proceedings of the Third International
Baltic Workshop on Databases and Information Systems, 1998, v.1, p.187-197.
50.Oksana Scegulnaja. Quantum real-time Turing machine. - Lecture
Notes in Computer Science, 2001, v.2138, p.412-415.
51.A.Spektors. Computer-aided language learning resources for the
Latvian language. - Proceedings of the Fifth International Conference
and Exhibition "Information Technologies and Telecommunications
in the Baltic States", 1998, p.234-237.
52.J.Viksna. Probabilistic inference of limiting functions with bounded
number of mind changes. - International Journal of Foundations of
Computer Science, 1996, v.7, No.3, p.187-208.
53.D.Zeps. Kuratowski theorem from below. - Acta Societatis Mathematicae
Latviensis, 2000, No.3, 69-70.
Recent
Projects
I
1.System
modeling tools GRADE. A joint
project with a/s DATI (Latvia) and Infologistik GmbH (Germany) (1991-2000).
2.Joint USA - Latvian Research in Inductive Inference. USA NSF grant
No.9421640 (1994-1998).
3.Trans-European Language Resources Infrastructure (TELRI). EU project
No.1202 (1995-1997).
4.Onomastica-Copernicus. EU project No.COP58 (1995-1997).
5.Software for Wireless Access Network (Wireless Local Loop). Contract
with DTR (Digital Trunked Radio) Inc. (1996-2000).
6.Universal Networking Language (UNL). UN University project (1997-1999).
7.Application Development for the Distributed Enterprise (ADDE). EU
4th Framework Programme ESPRIT project No. 23287 (1997-2000).
8.ML 2000 - Machine Learning and Effective Inductive Inference. The
Swedish Institute project (1997-2000).
9.European Language Activity Network. EU project No.6295 (1998-1999).
10.Trans-European Language Resources Infrastructure (TELRI-II). EU
project No.PL97-7085 (1998-2001).
11.Quantum Algorithms and Information Processing. EU 5th Framework
Programme project No.IST-1999-11234 (2000-2002).
12.Information Dissemination and European Awareness Launch for the
IST Programme (IDEALIST). EU 5th Framework Programme project No.IST-1999-14184
(2000-2003).
13.Cisco Networking Academy. Contract No. 35973 with Cisco Systems
Inc. (started in 2000).
14.Interactive Graph Layout and Diagramming. Contract with Tom Sawyer
Software Corporation (started in 2000).
15.Software for Real Time Application. Contract with TELDS SYSTEMS
(started in 2000).
16.CULTIVATE - CEE Cultural Heritage Applications Network Part. EU
5th Framework Programme project No.IST-1999-13472 (2001-2003).
II
System
analysis, requirement specification, software development. According
to contracts with organizations and ministries.
|
