English Version

Dr. Marius-Nicolae Popescu







Marius Popescu, Lecturer at the University of Bucharest, Department of Computer Science, since 2004. Ph.D. in 2004 with the thesis "Machine Learning Applied in Natural Language Processing".

My domains of interest are: artificial intelligence, machine learning, computational linguistics, information retrieval, authorship identification, computer vision. My achievements in these fields include: a new method for automatic word sense disambiguation that was awarded third prize at the Senseval 3 evaluation exercises for word sense disambiguation organized by ACL-SIGLEX in 2004; coauthor of the method ENCOPLOT for plagiarism detection which won the first international competition in automatic plagiarism detection in 2009, followed by ranking 4th at PAN@CLEF 2010 and 2nd at PAN@CLEF 2011; a method for authorship analysis which obtained the best results in the author identification task at PAN@CLEF 2012; Co-author of learning systems that ranked on 3rd place in the Native Language Identification Shared Task of the BEA 8 Workshop of NAACL 2013, and on 4th place in the Facial Expression Recognition Challenge of the WREPL Workshop of ICML 2013.

Publications:
  1. Marius Popescu. Training Strategies in Learning to Play Games. Annals of the University of Bucharest, Mathematics - Computer Science, Special Issue, Proceedings of Annual Meeting of Faculty of Mathematics, 28-30 nov. 1996, p. 147-163.
  2. Marius Popescu. Maximum Entropy Models for Dependency Parsing. Annals of the University of Bucharest, COMPUTER SCIENCE, an XLVI, 1997, p. 115-122.
  3. Florentina Hristea, Marius Popescu. A Word Grammar Approach to Syntactic Analysis with Special Reference to Romanian. Annals of the University of Bucharest, Mathematics - Computer Science, Special Issue, Proceedings of the Annual Meeting of the Faculty of Mathematics, 8-9 May 1998, p. 101-111.
  4. Florentina Hristea, Marius Popescu. A Dependency Grammar Approach to Syntactic Analysis with Special Reference to Romanian. In: Building Awareness in Language Technology (eds. Florentina Hristea and Marius Popescu). Editura Universitatii din Bucuresti, 2003, p. 9-16.
  5. Marius Popescu. Dependency Grammar Annotator. In: Building Awareness in Language Technology (eds. Florentina Hristea and Marius Popescu). Editura Universitatii din Bucuresti, 2003, p.17-34.
  6. Marius Popescu. Regularized Least-Squares Classification for Word Sense Disambiguation. Proceedings of The Third International Workshop on the Evaluation of Systems for the Semantic Analysis of Text (Senseval-3). ACL-2004 Workshop. Barcelona: Jul 2004, p. 209-212.
  7. Florentina Hristea, Marius Popescu. On Some Romanian Text Summarization Contributions. Revue Roumaine de Linguistique, VOL. LI, Issues 3-4, 2006, p. 507-509.
  8. Marius Popescu, Liviu P. Dinu Kernel Methods and String Kernels for Authorship Identification: The Federalist Papers Case. Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP-07). Borovets, Bulgaria. 2007.
  9. Marius Popescu, Liviu P. Dinu. Rank Distance as a Stylistic Similarity. Proceedings of the 22nd International Conference on Computational Linguistics (Coling 2008). Manchester, UK. 2008.
  10. Liviu P. Dinu, Marius Popescu. A Multi-Criteria Decision Method Based on Rank Distance. Fundamenta Informaticae. 86(1-2), 2008.
  11. Florentina Hristea, Marius Popescu, Monica Dumitrescu. Performing word sense disambiguation at the border between unsupervised and knowledge-based techniques. Artificial Intelligence Review. 30(1-4), 2008.
  12. Florentina Hristea, Marius Popescu. Adjective Sense Disambiguation at the Border Between Unsupervised and Knowledge-Based Techniques. Fundamenta Informaticae. 91(3-4), 2009.
  13. Liviu P. Dinu, Marius Popescu. A Language Independent Kernel Method for Classifying Texts with Disputed Paternity. Proceedings of the XIII International Conference Applied Stochastic Models and Data Analysis (ASMDA 2009). Vilnius, Lithuania. 2009.
  14. Vivi Nastase, Marius Popescu. What.s in a name? In some languages, grammatical gender. Proceedings of the EMNLP 2009 Conference on Empirical Methods in Natural Language Processing. Singapore. 2009.
  15. Marius Popescu, Liviu P. Dinu. Comparing Statistical Similarity Measures for Stylistic Multivariate Analysis. Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP-09). Borovets, Bulgaria. 2009.
  16. Cristian Grozea, Christian Gehl, Marius Popescu. ENCOPLOT: Pairwise Sequence Matching in Linear Time Applied to Plagiarism Detection. Proceedings of the 3rd PAN Workshop: Uncovering Plagiarism, Authorship and Social Software Misuse. San Sebastian, Spain. 2009.
  17. Liviu P. Dinu, Marius Popescu. Ordinal measures in authorship identification. Proceedings of the 3rd PAN Workshop: Uncovering Plagiarism, Authorship and Social Software Misuse. San Sebastian, Spain. 2009.
  18. Cristian Grozea, Marius Popescu. Who's the Thief? Determining the Direction of Plagiarism. 11th International Conference on Intelligent Text Processing and Computational Linguistics, CICLing-2010: March 21-27, 2010, Iasi, Romania. Proceedings. Lecture Notes in Computer Science, 6008, Springer, 2010.
  19. Cristian Grozea, Marius Popescu. Encoplot - Performance in the Second International Plagiarism Detection Challenge. Lab Report for PAN at CLEF 2010 . CLEF (Notebook Papers/LABs/Workshops) 2010.
  20. Marius Popescu, Florentina Hristea. State of the art versus classical clustering for unsupervised word sense disambiguation. Artificial Intelligence Review. 35(3), 2011.
  21. Cristian Grozea, Marius Popescu. The Encoplot Similarity Measure for Automatic Detection of Plagiarism. - Notebook for PAN at CLEF 2011 . CLEF (Notebook Papers/LABs/Workshops) 2011.
  22. Marius Popescu. Studying Translationese at the Character Level. Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP 2011). Hissar, Bulgaria. 2011.
  23. Marius Popescu, Cristian Grozea. Kernel Methods and String Kernels for Authorship Analysis. CLEF (Online Working Notes/Labs/Workshop) 2012.
  24. Cristian Grozea, Marius Popescu. Encoplot - Tuned for High Recall (also Proposing a New Plagiarism Detection Score). CLEF (Online Working Notes/Labs/Workshop) 2012.
  25. Dinu, Liviu Petrisor, Radu-Tudor Ionescu, Marius Popescu. Local patch dissimilarity for images. Neural Information Processing. Springer Berlin Heidelberg, 2012.
  26. Popescu, Marius, Radu Tudor Ionescu. The Story of the Characters, the DNA and the Native Language. NAACL/HLT 2013 (2013): 270.
  27. Ionescu, Radu Tudor, Marius Popescu, Cristian Grozea. Local Learning to Improve Bag of Visual Words Model for Facial Expression Recognition. ICML 2013 Workshop on Representation Learning, Atlanta, Georgia, USA, 2013.


BOOKS; EDITOR:

Florentina Hristea, Marius Popescu (eds.). Building Awareness in Language Technology. Editura Universitatii din Bucuresti, 2003 (478 pag.).




Programe de cercetare, Granturi - Fac. de Mat. si Inf.

 

 Modelling uncertainty in non-classical logics 

Contract nr. 200/01/10/2015 finantat de CNCS-UEFISCDI

 "Grupuri, grupuri cuantice, coringuri si teoria reprezentarii" 

 "Groups, quantum groups, corings and representation theory" 

Programul de cercetare face parte din Planul National de Cercetare, Dezvoltare si Inovare PN II, Programul "IDEI".

 "Geometrie de contact, complexa si quaternionica
pe varietati conforme si Riemann" 

Tema face parte din Programul National de Cercetare de Excelenta (CEEX-2006), Modulul 1.

Proiecte de cercetare pentru tineri cercetatori

 "Probleme recente in teoria subvarietatilor" 

Tema face parte din Programul National de Cercetare de Excelenta (CEEX) Modulul II - Proiecte de Dezvoltare a Resurselor Umane pentru Cercetare.

Tehnici statistice de Data Mining pentru

 "Software bazat pe metode avansate de decizie pentru o agricultura durabila" 

Tema face parte din Programul National de Cercetare de Excelenta (CEEX) Modulul I - Proiecte de cercetare-dezvoltare complexe.

Instrumente pentru promovarea domeniului

 "Luarea deciziei multi-atribut" 

Tema face parte din Programul National de Cercetare de Excelenta (CEEX) si este o componenta a Proiectului Instrumente la nivel de excelenta pentru promovarea domeniului LUAREA DECIZIEI MULTI-ATRIBUT.

 "Tehnici matriceale si operatoriale in analiza functionala" 

COD CNCSIS: 1905, contract nr. 538/2009

 "Aplicatii ale unor clase speciale de inegalitati variationale si probleme de echilibru in finante si nanoindentare" 

Programul de cercetare face parte din Planul National de Cercetare, Dezvoltare si Inovare PN II, Programul "IDEI",.cod ID, nr.112/01.10.2007

 "Coinele, Algebre Hopf si Categorii Braided Monoidale" 

Programul de cercetare face parte din Planul National de Cercetare, Dezvoltare si Inovare PN II, COD IDEI 1005.

Scan Statistics si Aplicatii

 "Scan Statistics si Aplicatii" 

Colaborarea se realizeaza intre Centrul de cercetari in Informatica medicala (CERIM) al Universitatii Lille2 si Colectivul de cercetare in Informatica al Facultatii de Matematica si Informatica, Universitatea din Bucuresti.

F.M.I.

Centru Regional de Informare in Balcani
pentru tehnologia limbajului

 Proiectul BALRIC-LING 

 "ACTIUNI, COACTIUNI SI GRADUARI PE ALGEBRE" 

Programul de cercetare face parte din Planul National de Cercetare, Dezvoltare si Inovare PN II, COD IDEI 1904.

 SUBMERSII SI SUBVARIETATI DE TIP CUATERNIONIC 

CNCSIS PN II IDEI nr. 525/2009, Cod CNCSIS ID_8

 Efecte plastice si vascoase in materiale ne-elastice supuse la deformari 

Programul de cercetare face parte din Planul National de Cercetare, Dezvoltare si Inovare PN II, COD IDEI ID_1248, contract nr. 567/2009.

 ALGEBRE HOPF, COOMOLOGIE CICLICA SI CATEGORII MONOIDALE 

CNCSIS PN II IDEI nr.560/2009, cod CNCSIS ID_69


2002 - 2015 -- Facultatea de Matematica si Informatica, Universitatea din Bucuresti
Str. Academiei nr. 14, sector 1, C.P. 010014, Bucuresti, Romania
Tel: (4-021) 314 2863, Fax: (4-021) 315 6990, secretariat  fmi.unibuc.ro