{ LAURENT KEVERS,.N+PERS }

Version française [ Work experience | Education | Papers and communications | Scientific activities | Personal interests | Contact ]

I'm a computational linguist and researcher in natural language processing (NLP). More specifically, my area of expertise is information extraction and retrieval as well as knowledge modelisation and management. These fields allowed me to address a wide variety of NLP problems as Named Entity Recognition and Extraction, Temporal Information Extraction, Text Categorization or Semantic access to textual documents. I invite you to visit this page as well as my Academia and Linked in profiles to learn some details of my résumé.

CV (French)   Résumé (English)

Knowledge and Technology Transfer Advisor
When: 2012-....
Where: ADRE / Technology Transfer Office - University of Namur
What: In the fields of Information and Communication Technology (ICT), and Social and Human Sciences (SHS).
Research : Projects
Stratego
When: 2007-2010
What: The main goal of the Stratego project consists in the development of tools to improve information retrieval through large scale textual databases. More precisely, this project focus on (i) automatic structuring of raw documents (e.g. digitalized documents) into XML documents compliant with a DTD or a XML schema, (ii) automatic classification of documents in predefined categories, and (iii) semi-automatic thesauri generation using specialized corpus (e.g. law texts).
Activities: Extraction of features based on linguistic methods for categorization : Multi Words Expressions (MWE) and Multi Words Units (MWU), Named Entities... Creation of an automatic categorization system for textual documents with regard to a defined set of categories (a thesaurus) through a linguistic method based on a resource automatically generated from the set of categories itself (from the thesaurus).
Who: IRIS (industrial partner), Cental (Centre for Natural Language Processing), IRIDIA (ULB) (Artificial Intelligence research laboratory of the University of Brussels), ISYS (UCL) (Information Systems Unit) et SIC (ULB) (Information and Communication Sciences Department, University of Brussels)
B-Ontology
When: 2005-2007
What: B-Ontology is a research project which aims to build a software prototype able to automatically extract and organize biographical information. This information is then usable by the press industry. The belgian press agency Belga broadcasts on a daily basis more than 250 press dispatches in two languages (french and dutch). This textual collection represents more or less 70,000 words per day (25 millions words per year) for each language. For a company like Belga, which defines itself as an information broker, it is a strategic issue to make the most of this information stream. The main use considered in this project is to manage information about people, but also about organizations or events related to these persons. The result is organized according to an ontology and memorized in an knowledge base.
Activities: Information Extraction (IE), Named Entities Recognition (NER), Knowledge bases and ontologies.
Who: Belga (industrial partner) and Cental (Centre for Natural Language Processing)
[moca], multimodal oral corpora administration
When: 2004-2006
What: Electronic management and use of large oral corpora.
Activities: Global system design, database engineering, web interface design.
Who: Valibel (data and design) and Cental (design and technical expertise)
Julibel
When: 2004
What: This database contains small excerpts from various sources (litterature, advertisments, oral corpora...), that can be used during french courses in order to illustrate various features of the language, as irony, connectives, stylistic devices...
Activities: Database and web interface redesign.
Who: Valibel (data) and Cental (technical expertise)
Research project in Greek lexicology (Projet de Recherche en Lexicologie Grecque, PRLG)
When: 2003-2005
What: Production of lemmatized concordances of Greek patristic and Byzantine authors. The project goal is to provide a comprehensive view of the Greek lexicon through the lemmatization of patristic and Byzantine texts.
Activities: Contribution to the evolution of a publishing process from "paper/microfiche" to a digital format : automated morpho-syntactic labelling and partial disambiguation of texts (C language and Unitex), lemmatization interface integrated into Unitex (Java), online concordances.
Who: Institut Orientaliste (data) and Cental (technical expertise)
2007 - 2011
Ph.D in Computational Linguistics (Doctorat en « langues et lettres »), Centre for Natural Language Processing (Cental), UCL, Belgium.
Thesis : ”Semantic access to databases of documents. Natural language processing with symbolic techniques for thematic indexation and temporal information extraction”.
Director : Cédrick Fairon.
1999 - 2002
Maîtrise (Bac+5) in Computer Science (graduated with high honours), University of Namur (FUNDP), Belgium.
Master thesis : L. Kevers and J.-B. Van Zuylen, Creation and use of digital audiovisual archives
Director : Jean-Marie Jacquet.
1996 - 1999
Candidature (Bac+2) in Economics, Social Sciences and Business Administration, Computer Science specialization, University of Namur (FUNDP), Belgium.
Kevers L. (2011), Accès sémantique aux bases de données documentaires. Techniques symboliques de traitement automatique du langage pour l'indexation thématique et l'extraction d'informations temporelles, thèse de doctorat en Langues et lettres, Université catholique de Louvain, 31-01-2011.
Text   Talk

(EN) ”Semantic access to databases of documents. Natural language processing with symbolic techniques for thematic indexation and temporal information extraction”.

Within our information society and economy, access to the growing bulk of electronic documents is a major concern. Surprisingly, information retrieval tools often only work on a classical search scheme, based on words. But, as a matter of fact, the relation between words and concepts can not always be reduced to a single link. A concept can be named by several expressions (« unemployed person » and « job-seeker »), and conversely one word can sometimes depict more than one concept (« carrot » as a vegetable or in geology). To cope with this difficulty, it might be interesting to represent documents in a concepts space in place of a words space.

In this thesis we are proposing 3 methods designed to complete the documents representation with semantic data : a (semi) automatic supervised classification method was set up to index documents, the specific processing of some information items, such as time, and finally a thematic and temporal indexing system that outlines interest and perspectives for applications within this multidimensional and semantic approach.

L. Kevers et B. Kindt (2005), Traitement automatisé de l'ambiguïté lexicale en grec ancien. Première approche par application de grammaires locales. In Lingvisticae Investigationes, 28: 235-254.
L. Kevers et B. Kindt (2004). "Vers un concordanceur-lemmatiseur en ligne du grec ancien". In L'Antiquité Classique 73: p. 203-213.
L. Kevers et J. Medori (2010), Symbolic classification methods for patient discharge summaries encoding into ICD, In: Advances in Natural Language Processing, 7th International Conference on NLP, IceTAL 2010, Reykjavik, August 16-18, 2010, Lecture Notes in Artificial Intelligence, 2010, p. 197-208.
L. Kevers, A. Mantrach, C. Fairon, H. Bersini, M. Saerens (2010), Classification supervisée hybride par motifs lexicaux étendus et classificateurs SVM, In: 10th International Conference on statistical analysis of textual data (JADT 2010), Rome, 9-11/06/2010, S. Bolasco, I. Chiari, L. Giuliano ed(s), Ed. Univ. di Lettere Economia Diritto, 2010, p. 105-117.
L. Kevers (2009), "Indexation semi-automatique de textes : thésaurus et transducteurs", In: Actes de la Sixième Conférence Francophone en Recherche d'Information et Applications (CORIA), Presqu'Île de Giens, p. 151-167.
L. Kevers et C. Fairon (2007), "Vers une base de connaissances biographiques : extraction d'information et ontologies". In: Revue des Nouvelles Technologies de l'Information - Série Extraction et gestion des connaissances. RNTI-E-9. Actes des 7èmes Journées Francophones Extraction et Gestion des Connaissances (EGC'2007, Namur). Cépaduès éditions. Vol. 1: 373-378.
L. Kevers (2006), "L'information biographique : modélisation, extraction et organisation en base de connaissances". In: P. Mertens, C. Fairon, A. Dister, P. Watrin (éds). Verbum ex machina. Actes de la 13e conférence sur le Traitement automatique des langues naturelles (TALN06). Presses universitaires de Louvain, Louvain-la-Neuve (Cahiers du Cental 2): 680-689.
Y. Bestgen, C. Fairon et L. Kevers (2004). "Un baromètre affectif effectif. Corpus de référence et méthode pour déterminer la valence affective de phrases". In: G. Purnelle, C. Fairon et A. Dister (eds). Le poids des mots. Actes des 7es Journées internationales d'Analyse statistique des Données Textuelles (JADT04). Presses universitaires de Louvain.
P. Gilles, L. Kevers et A.-C. Simon (2006), MOCA: un système de gestion et d'annotation de données orales. 3ème Rencontre Fribourgeoise de la Linguistique sur Corpus.
L. Kevers et B. Kindt (2005), Adaptation des ressources d'Unitex au traitement du grec ancien. 24ème conférence internationale grammaires et lexiques comparés, Liverpool.
L. Kevers et B. Kindt (2004), Traitement automatisé de l'ambiguïté lexicale en grec ancien. Première approche par application de grammaires locales. 23ème conférence internationale grammaires et lexiques comparés, Deauville.
Seminars and other presentations

2007 – Cental : "Processing of temporal expressions in press dispatches".

2005 – ATALA : "Automated processing of ancient Greek and of non-classic Arabic" (with B. Kindt and L. Tuerlinckx).

2003 – LADL : "Lemmatization of Greek patristic and byzantine sources for a lexical description of ancient Greek" (with B. Kindt).

Other scientific activities

Travelling is for me the unique opporutnity to live simultaneously a few passions. Hiking and trekking across unforgettable sceneries, meet amazing people and cultures, and try to express all these feelings through photography. I had the chance to give free expression to these passions during a sabbatical year (2011-2012) wich allowed me to visit 15 countries all over the world.

I also traveld and walked through these tracks :
- the Alps (Massif des Ecrins)
- the Cévennes (Stevenson track)
- the Pyrénées (Mont Perdu)
- Corsica (GR20)
- Australia (Overland Track, Tasmania)
- Iceland (Laugavegur, from Landmanalaugar to Þórsmörk)

Please find my contact detail into my Résumé.