A5080107492- Natural Language Processing Techniques
- Topic Modeling
- Text Readability and Simplification
- linguistics and terminology studies
- French Language Learning Methods
- Lexicography and Language Studies
- Speech and dialogue systems
- Linguistics and Discourse Analysis
- Linguistic research and analysis
- Linguistics and language evolution
- Translation Studies and Practices
- Medieval European Literature and History
- Biomedical Text Mining and Ontologies
- Semantic Web and Ontologies
- Expert finding and Q&A systems
- Historical Linguistics and Language Studies
- Authorship Attribution and Profiling
- Linguistic and Sociocultural Studies
- Advanced Text Analysis Techniques
- Spanish Linguistics and Language Studies
- Second Language Acquisition and Learning
- Wikis in Education and Collaboration
- Linguistic Variation and Morphology
- Linguistics, Language Diversity, and Identity
- Music and Audio Processing
Université de Strasbourg
2012-2025
Charles University
2024
University of Potsdam
2024
University of Washington
2024
Freie Universität Berlin
2024
Academy of Performing Arts in Prague
2024
University of Stuttgart
2024
University of Cologne
2024
Technical University of Darmstadt
2008-2021
Bar-Ilan University
2021
Bien que le dagur et l’alsacien représentent deux familles de langues typologiquement éloignées, ils partagent plusieurs similitudes : les sont en danger, n’ont pas système orthographique unifié, ont peu corpus numériques disponibles. Compte tenu ces défis, l’objectif principal cet article est comparer bruit dans son impact sur l’annotation l’étiquetage des parties du discours (POS). Nous discutons d’abord stratégies qui peuvent être utilisées pour réduire dû aux incohérences orthographiques...
Information overload is a well-known problem which can be particularly detrimental to learners. In this paper, we propose method support learners in the information seeking process consists answering their questions by retrieving question paraphrases and corresponding answers from social Q&A sites. Given novelty of kind data, it crucial get better understanding how sites automatically analysed retrieved. We discuss evaluate several pre-processing strategies similarity metrics, using new...
Monolingual translation probabilities have recently been introduced in retrieval models to solve the lexical gap problem. They can be obtained by training statistical on parallel monolingual corpora, such as question-answer pairs, where answers act "source" language and questions "target" language. In this paper, we propose use a dataset definitions glosses provided for same term different semantic resources. We compare built from resources with two other kinds of datasets: manually-tagged...
This paper describes the approaches authors developed while participating in i2b2/VA 2010 challenge to automatically extract medical concepts and annotate assertions on relations between concepts.The authors'approaches rely both rule-based machine-learning methods. Natural language processing is used features from input texts; these are then authors' approaches. The Conditional Random Fields for concept extraction, Support Vector Machines assertion relation annotation. Depending task, tested...
This paper presents a method for the syntactic simplification of French texts. Syntactic aims at making texts easier to understand by simplifying complex structures that hinder reading. Our approach is based on study two parallel corpora (encyclopaedia articles and tales). It identify linguistic phenomena involved in manual organise them within typology. We then propose system relies this typology generate simplified sentences. The module starts generating all possible variants before...
This article describes a question generation system for French. The transformation of declarative sentences into questions relies on two different syntactic parsers and named entity recognition tools. makes it possible to further diversify the generated possibly alleviate problems inherent analysis also generates reformulations based variations in words, inducing answers with granularities, nominalisations action verbs. We evaluate extracted from corpora: corpus newspaper articles used CLEF...
In this paper, we present the system have developed for participating in second task of i2b2/VA 2011 challenge dedicated to emotion detection clinical records. On official evaluation, ranked 6th out 26 participants. Our best configuration, based upon a combination both machine-learning approach and manually-defined transducers, obtained 0.5383 global F-measure, while distribution other participants' results is characterized by mean = 0.4875, stdev 0.0742, min 0.2967, max 0.6139, median...
Morphology is a key component for many Language Technology applications. However, morphological relations, especially those relying on the derivation and compounding processes, are often addressed in superficial manner. In this article, we focus assessing relevance of deep motivated knowledge Natural Processing We first describe an annotation experiment whose goal to evaluate role morphology one task, namely Question Answering (QA). then highlight kind linguistic that necessary particular...
Morphologically complex terms composed from Greek or Latin elements are frequent in scientific and technical texts. Word forming units thus relevant cues for the identification of domain-specific This article describes a method automatic extraction relying on detection classical prefixes word-initial combining forms. Word-forming identified using regular expression. The system then extracts by selecting words which either begin coalesce with these elements. Next, grouped families displayed...