Liviu CIORTUZ: Short Curriculum Vitae
PhD in Computer Science, 1996,
Thesis: DF - A Feature Constraint Concurrent System with Application to Natural Language Processing.
Supervised by Prof. Jean-Paul Delahaye.
MSc in Mathematics and Computer Science, 1985,
Thesis: Stochastic Variables and Martingal Processes.
Associate Professor at the Computer Science Department ,
Machine Learning (2003-today), undergraduate level, and MSc level (2009-today)
Bioinformatics (2006-today), undergraduate level, and MSc level (2005-2009)
· I am interested in applying Machine Learning algorithms and techniques to the area of Bioinformatics and Natural Language Processing.
· Prior to this, I worked on applying implementation techniques from (constraint) Logic Programming to the compilation of (parsing with) unification grammars.
· Among the different systems I implemented, the most important one is LIGHT, a compiler for large-scale HPSG-like grammars.
system is documented by a number of papers that I published mainly during
2002-2004. LIGHT is capable of running the LinGO ERG grammar for English
developed at CSLI,
malware detection (BitDefender),
computer games (McGnuGO).
LIGHT - a compiler for head-driven parsing with unification grammars (C)
ilpLIGHT - a machine learning system for typed-unification grammars (C), based on adapting Inductive Logic Programming to feature-constraint (OSF) logic
LOGOS - an object-oriented framework to implement logic languages (C++), in particular Proplog, Datalog and Prolog
DF - a feature constraint concurrent language (Oz), in the framework of F-logic
CHALLENGER - a prototype Machine Translation system from English into Romanian (C/Oz)
HPSG RoCK - a parser for a HPSG subset of Romanian (Oz)
… (not yet named) – a system for malware detection based on perceptrons and SVMs, currently used by the BitDefender antivirus engine to reduce the number of false alarms; co-designers and implementors: Dragos Gavrilut, Mihai Cimpoesu, Dan Anton
yasMiR – a SVM-based system for the identification of microRNAs; implementors: Daniel Pasaila (main author), Andrei Sucila, Irina Mohorianu, and Stefan Pantiru
McGnuGO – a program for playing GO, using the Monte Carlo machine learning method; main author and (single) implementor: Florin Chelaru
November 2002 - September 2003:
associate at the Computer Science Department of the
January 2001 - September 2002:
Research associate at the Computer Science Department of the University of York, UK. Worked on the EPSRC project ``Machine Learning for Natural Language Processing in a Computational Logic framework'': Inductive Logic Programming for automate learning of typed-unification grammars.
June 1997 - January 2001:
researcher at the Language Technology Lab of DFKI, the
October 1990 - June 1997:
Assistant and then lecturer at the Faculty of Computer Science University of Iasi, Romania.
Given courses and seminaries:
Statistical Natural Language Processing
(2004-2006), diploma level, and MS level (2003-2006)
Machine Learning (1992-1994), diploma level
Natural Language Processing (1990-1994), seminaries, diploma level
Machine Translation (compact course, October 2002), MS level
Concurrent Constraint Programming (1996-1997), MS level
Database Systems (1990-1992), seminaries, diploma level
Object-oriented Software Engineering (1992-1994), diploma level
Mathematical Logic and Logic Programming (1991-1993), DEng level
November 1994 - December 1996:
part-time PhD student at
October 1985 - September 1990:
Research assitant at
the Institute for Computer Science of the
C/C++, Prolog, Oz; some experience in SQL, Lisp, and Perl.
Last updated: 12th February, 2010