Audible Inc.

E-mail:
dstefanescu@gmail.com
stefanes@audible.com

Google Scholar
LinkedIn
ResearchGate

Dan Ştefănescu

Academic Background

2013 - 2014, Department of Computer Science, University of Memphis, Memphis, TN, USA
Postdoc. on Intelligent Tutoring Systems

2005 - 2010, Research Institute for Artificial Intelligence, Romanian Academy, Bucharest, Romania
Ph.D., thesis: Intelligent Information Mining from Multilingual Corpora

2002 - 2004, Faculty of Computer Science, "Al. I. Cuza" University, Iaşi, Romania
M.Sc. in Computational Linguistics, dissertation: Automatic Recognition of Protein Names in Biomedical Texts

1998 - 2002, Faculty of Computer Science, "Al. I. Cuza" University, Iaşi, Romania
B.Sc., thesis: Searching the World Wide Web using Natural Language Queries

Work Experience

since June 2015: Data Scientist
Audible Inc.

2014 - 2015: Senior Researcher and Software Developer
Vantage Labs

2013 - 2014: Post-Doctoral Research Fellow
Department of Computer Science @ University of Memphis
DeepTutor Project @ Institute for Intelligent Systems, FedEx Institute of Technology

2004 - 2012: Research Assistant / Researcher / Senior Researcher III
Research Institute for Artificial Intelligence @ Romanian Academy, Bucharest, Romania
Natural Language Processing Group

Teaching:

Statistical Machine Translation @ Faculty of Computer Science, Al. I. Cuza Iasi (2010)
Machine Translation @ EurolanJunior Faculty of Computer Science, Al. I. Cuza Iasi (2009)

Languages

Romanian (native), English (fluent), French (some)

Interests

Natural Language Processing

  • Semantic Similarity / Relatedness
  • Word Sense Disambiguation
  • Question Answering
  • Language Modeling
  • Context Sensitive Spell Checking
  • Dialogue / Tutoring Systems
  • Parallel Data Extraction from Comparable Corpora
  • Automatic Terminology Indentification
  • Statistical Information Mining from Multilingual Corpora
  • Statistical Machine Translation

Machine Learning

  • Regression Models
  • Markov Models
  • Maximum Entropy Models
  • Latent Semantic Analysis
  • Artificial Neural Networks & Deep Learning
  • Support Vector Machines
  • Genetic Algorithms

Always interested in learning

Hobbies

Results in competitions

  • 1st place (+ DeepTutor team @ University of Memphis) - Interpretable Semantic Textual Similarity task @ SemEval 2015
  • 4th place (+ Radu Ion & Tiberiu Boroş @ RACAI) - Microsoft Speller Challenge 2011
  • 1st place (as member of RACAI NLP team) - Question-Answering System for JRC-Acquis track, CLEF 2009
  • 1st place (as member of RACAI NLP team) - Question-Answering System for Ro-Ro Wiki track, CLEF 2007
  • 1st place (as member of RACAI NLP team) - Lexical Unit Alignment Competition for English-Romanian pair of languages, ACL 2005, USA
  • 3rd place (+ Alexandru Ceauşu @ RACAI) - regional phase (Romania and Republic of Moldova) of the Microsoft Imagine Cup 2005
  • 2nd place - EconomMIX Communication Session Contest, FEAA, Univ. "Al. I. Cuza", Iaşi, Romania, 2003

Applications and resources developed

    Obsolete:

  • Named Entity Recognizer (C#)
  • System for web queries alteration (+ Radu Ion & Tiberiu Boroş): one of the Microsoft Speller Challenge Winners (C#)
  • DiacriticsROi - Application for Romanian diacritics insertion (C#)
  • WSD System based on Lexical Chains and Graph Context Formalization (C#)
  • MTKit (Word Aligner + editor) (+ Alexandru Ceauşu) (C#)
  • SVM Based Sentence Aligner (+ Alexandru Ceauşu) (C#)
  • QA System on JRC-Acquis, an effort of the entire NLP team of RACAI (C#)
  • Maximum Entropy Question Classifier (C#)
  • Net Search Library (C#)
  • Multilingual Thesauri Aligner (C#)
  • Romanian Hyphenator (Active Perl)

All Publications

Articles in Conference proceedings, Journals and Magazines

    2009

  • Extracting Collocations in Contexts
  •    Amalia Todiraşcu, Christopher Gledhill, Dan Ştefănescu
       In Human Language Technology. Challenges of the Information Society, Lecture Notes in Computer Science (LNCS), Volume 5603, pp 336-349
  • RACAI’s QA System at the Romanian-Romanian QA@CLEF2008 Main Task
  •    Radu Ion, Dan Ştefănescu, Alexandru Ceauşu, Dan Tufiş
       In Evaluating Systems for Multilingual and Multimodal Information Access, Lecture Notes in Computer Science (LNCS), Volume 5706, pp 393-400
  • Unsupervised Word Sense Disambiguation with Lexical Chains and Graph-based Context Formalization
  •    Radu Ion, Dan Ştefănescu
       In Proceedings of the 4th Language & Technology Conference (LTC), November 6-8, Poznan, Poland
  • A Trainable Multi-factored QA System
  •    Radu Ion, Dan Ştefănescu, Alexandru Ceauşu, Dan Tufiş, Elena Irimia, Verginica Barbu-Mititelu
       In Proceedings of CLEF2009 Workshop, Sept. 30 - Oct. 2, Corfu, Greece
  • CONAN – Detecția posibilelor conotații ale unui text (romanian)
  •    Dan Ştefănescu, Dan Tufiş
       Proceedings of the 4th International Conference "Linguistic Resources and Tools for Processing of the Romanian Language" (ConsILR), pp 141-150, Iaşi, Romania
  • Resurse lingvistice pentru un sistem de întrebare-răspuns pentru limba română (romanian)
  •    Verginica Barbu Mititelu, Alexandru Ceauşu, Radu Ion, Elena Irimia, Dan Ştefănescu, Dan Tufiş
       In Revista Româna de Interactiune Om-Calculator 2, pp 1-17, Bucharest, Romania
    2008

  • Important Practical Aspects of an Open-domain QA System Prototyping
  •    Radu Ion, Dan Ştefănescu, Alexandru Ceauşu
       In Proceedings of the Romanian Academy, Series A, p 6, The Publishing House of the Romanian Academy, Bucharest, Romania
  • RACAI's Question Answering System at QA@CLEF 2007
  •    Dan Tufiş, Dan Ştefănescu, Radu Ion, Alexandru Ceauşu
       In Advances in Multilingual and Multimodal Information Retrieval (CLEF 2007), Volume 5152 of Lecture Notes in Computer Science (LNCS), pp 3284-3291
  • RACAI's Linguistic Web Services
  •    Dan Tufiş, Radu Ion, Alexandru Ceauşu, Dan Ştefănescu
       In Proceedings of the 6th Language Resources and Evaluation Conference (LREC), May 28-30, Marrakech, Morocco
  • Extraction de collocations monolingues et bilingues: application a la traduction (french)
  •    Dan Ştefănescu, Alexandru Ceauşu, Radu Ion, Amalia Todiraşcu, Ulrich Heid, Christopher Gledhill, François Rousselot
       In Proceedings of the Latin Union Conference, February 28-29, Bucharest, Romania. ISBN 978-9-291220-37-3
  • RACAI's QA System at the Romanian-Romanian Multiple Language Question Answering (QA@CLEF2008) Main Task
  •    Radu Ion, Dan Ştefănescu, Alexandru Ceauşu, Dan Tufiş
       In Working Notes for the CLEF 2008 Workshop, p 10, Aarhus, Denmark
  • Vers un dictionnaire de collocations multilingue (french)
  •    Amalia Todiraşcu, Ulrich Heid, Dan Ştefănescu, Dan Tufiş, Christopher Gledhill, Marion Weller, Francois Rousselot
       In Cahier de Linguistique, 33/1, pp 161-186, Louvain. ISSN 0771-6524
  • ROTEL: Linguistic Web Services
  •    Dan Tufiş, Radu Ion, Alexandru Ceauşu, Dan Ştefănescu
       In Proceedings of the Excellence Research – A Way to Innovation (CEEX), pp 29-1 – 29-6, Braşov, Romania
  • A Mix Approach to Extracting and Classifying Verb+Noun Constructions
  •    Amalia Todiraşcu, Dan Tufiş, Ulrich Heid, Christopher Gledhill, Dan Ştefănescu, Marion Weller, François Rousselot
       In Proceedings of the 6th Language Resources and Evaluation Conference (LREC), May 28-30, Marrakech, Morocco
  • Romanian Wordnet: Current State, New Applications and Prospects
  •    Dan Tufiş, Radu Ion, Luigi Bozianu, Alexandru Ceauşu, Dan Ştefănescu
       In Proceedings of the 4th Global WordNet Conference (GWC), pp 441-452, Szeged, Hungary
    2007

  • Cross-Lingual Romanian to English Question Answering at CLEF 2006
  •    Georgiana Puşcaşu, Adrian Iftene, Ionut Pistol, Diana Trandabăţ, Dan Tufiş, Alexandru Ceauşu, Dan Ştefănescu, Radu Ion, Constantin Orăşan, Iustin Dornescu, Alex Moruz, Dan Cristea
       In Evaluation of Multilingual and Multi-modal Information Retrieval, Lecture Notes in Computer Science (LNCS), Volume 4730, pp 385-394
  • Extracting Collocations in Context
  •    Amalia Todiraşcu, Christopher Gledhill, Dan Ştefănescu
       In Proceedings of 3rd Language & Technology Conference (LTC), October 5-7, Poznan, Poland
  • Extracting Collocations in Context: the case of Romanian VN constructions
  •    Amalia Todiraşcu, Christopher Gledhill, Dan Ştefănescu
       In Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP), September 27-29, Borovets, Bulgaria
  • RACAI's Question Answering System at QA@CLEF 2007
  •    Dan Tufiş, Dan Ştefănescu, Radu Ion, Alexandru Ceauşu
       In Proceedings of the CLEF2007 Workshop, September 19-21, Budapest, Hungary
  • Un sistem de extragere a colocațiilor (romanian)
  •    Amalia Todiraşcu, Dan Ştefănescu, Christopher Gledhill
       In Proceedings of the 3rd International Conference "Linguistic Resources and Tools for Processing of the Romanian Language" (ConsILR), Iaşi, Romania
  • Servicii web lingvistice ale ICIA (romanian)
  •    Dan Tufiş, Radu Ion, Alexandru Ceauşu, Dan Ştefănescu
       In Proceedings of the 3rd International Conference "Linguistic Resources and Tools for Processing of the Romanian Language" (ConsILR), Iaşi, Romania
    2006

  • Aligning Multilingual Thesauri
  •    Dan Ştefănescu, Dan Tufiş
       In Proceedings of the 5th Language Resources and Evaluation Conference (LREC), May 22-28, Genoa, Italy
  • Improved Lexical Alignment by Combining Multiple Reified Alignments
  •    Dan Tufiş, Radu Ion, Alexandru Ceauşu, Dan Ştefănescu
       In Proceedings of the 11th Conference of the European Chapter of the Association for Computational Linguistics (EACL), April 3-7, Trento, Italy
  • Acquis Communautaire Sentence Alignment using Support Vector Machines
  •    Alexandru Ceauşu, Dan Ştefănescu, Dan Tufiş
       In Proceedings of the 5th Language Resources and Evaluation Conference (LREC), May 22-28, Genoa, Italy
  • Identificarea și extragerea automată a colocațiilor din texte (romanian)
  •    Dan Ştefănescu, Dan Tufiş, Elena Irimia
       In Proceedings of the 2nd International Conference "Linguistic Resources and Tools for Processing of the Romanian Language" (ConsILR), Nov. 3, Bucharest, Romania
  • Developing a Question Answering System for the Romanian-English Track at CLEF 2006
  •    Georgiana Puşcaşu, Adrian Iftene, Ionut Pistol, Diana Trandabăţ, Dan Tufiş, Alexandru Ceauşu, Dan Ştefănescu, Radu Ion, Constantin Orăşan, Iustin Dornescu, Alex Moruz, Dan Cristea
       In Proceedings of CLEF 2006 Workshop, September 20-22, Alicante, Spain
  • Sentence and word alignment using Support Vector Machines
  •    Alexandru Ceauşu, Dan Tufiş, Dan Ştefănescu
       In Proceedings of IST - Multidisciplinary approaches, Bucharest, Romania
  • Aligning Unequal Multilingual Thesauri
  •    Dan Ştefănescu
       In Proceedings of Central European Student Conference in Linguistics (CESCL), Budapest, Hungary
  • Resources, Tools and Algorithms for the Semantic Web
  •    Dan Tufiş, Radu Ion, Elena Irimia, Verginica Barbu Mititelu, Alexandru Ceauşu, Dan Ştefănescu, Luigi Bozianu, Cătălin Mihăilă
       In Proceedings of IST - Multidisciplinary approaches, Bucharest, Romania
    2005

  • Combined Word Alignments
  •    Dan Tufiş, Radu Ion, Alexandru Ceauşu, Dan Ştefănescu
       In Proceedings of the ACL Workshop on "Building and Using Parallel Corpora: Data-driven Machine Translation and Beyond", 29-30 June 2005, Ann Arbor, Michigan, USA
  • An integrated platform for high-accuracy word alignment
  •    Dan Tufiş, Alexandru Ceauşu, Radu Ion, Dan Ştefănescu
       In Proceedings of JRC Enlargement and Integration Workshop: Exploiting parallel corpora in up to 20 languages, Sep. 26-27, Arona, Italy
  • Word Senses and Cross-lingual Word Sense Disambiguation
  •    Dan Tufiş, Radu Ion, Dan Ştefănescu, Alexandru Ceauşu
       In Proceedings of EUROLAN 2005, July 25 - Aug. 6, Cluj, Romania