Idiap Research Institute, Martigny, Switzerland
Since February 2021, I am leading the Computation, Cognition & Language group at Idiap in Martigny. I am affiliated with the University of Malta where I have been senior lecturer in Human Language Technology since 2014. Before that, I was junior professor at the Institute for Natural Language Processing (IMS), University of Stuttgart, where I lead a research group (see below) in the framework of the SFB 732, a collaborative research centre that brings computational linguists and linguists together. I remained affiliated with the IMS as project co-leader till the end of the SFB732 and am now an external collaborator in the DFG project Zero-derived nouns and deverbal nominalization: An empirically-oriented perspective (2018-2021) by Gianina Iordachioaia. I did a post-doc (maître-assistante) at the University of Geneva working in the field of cross-lingual transfer of semantic role labelling as part of the CLASSiC project. I earned my PhD from the University of Groningen (Department of Humanities Computing), where I worked on vector-based word representations for medical question answering within the Alfa-Informatica group. I worked at ISSCO/TIM-ETI (University of Geneva) from 2002 until 2003. I worked in industry for one year at Systran Translation Systems in 2001-2002. Before that I did the M.Phil Computer Speech and Language Processing at the University of Cambridge. The M.Phil has now been renamed into Computer Speech, Text and Internet Technology.
I have been working on the following subjects: computational linguistics, cross-lingual natural language processing (nlp), distributional semantics, vector-based word representations, text mining, (medical) terminology extraction, computational creativity, (medical) question answering, semantic role labelling, low-resource languages, diachronic computational linguistics.
Opportunities for PhD students and postdocs at Idiap in Martigny, Switzerland:
If you are from Valais (having completed your primary and secondary schools in Valais), and subsequently followed EPFL-equivalent studies, you can apply for a Valais-Wallis Ambition initiative for PhDs and Postdocs. Appointments as a PhD student are typically for 4 years, conditional to successful progress, and should lead to a PhD dissertation, granted by the “Ecole Polytechnique Fédérale de Lausanne (EPFL), while conducting your research at Idiap in Martigny. Postdoc appointments are typically for one year, with possibilities of renewal for a maximum of 4 years, conditional to performance. Contact me for more information.
If you are a post-doc from abroad with an excellent track record and would like to conduct research on NLP at Idiap, contact me to discuss opportunities.
I was a DSI fellow at the University of Zurich from October 2019-January 2020
I was a visiting academic at the Division of Information and Communication Sciences of Macquarie University, Sydney from January till March 2007
H2020-WIDESPREAD (2021–2023, main applicant/coordinator)
Bridging the technology gap: Integrating Malta into European Research and Innovation efforts for AI-based language technologies.
UPSKILLS (UPgrading the SKIlls of Linguistics and Language Students)
EU Erasmus+ (2021-2023, main applicant/coordinator)
The central goal of this project is to tackle the identified skills gaps and mismatches in linguistics and language students through supporting the development of innovative materials that better meet the learning outcomes needed in the current job market, and this in collaboration with language technology companies.
MUFINS (NLP-Driven, MUltilingual FInancial News and content Search)
Malta Enterprise Research grant (2020-2022, co-applicant)
This project aims to investigate transfer learning techniques to resolve a variety of NLP tasks in multiple languages, within the financial
and news domains. This project is led in collaboration with CityFalcon Trading Ltd.
Post-doc: Marc Tanti
MASRI (Maltese Automatic Speech Recognition, https://www.um.edu.mt/projects/masri/)
UM Research Fund (2018-2020, co-applicant)
This project aims to build a automatic speech recogniser for Maltese, a low-resource language, using various bootstrapping approaches.
SFB 732 project D11: A crosslingual approach to the analysis of compound nouns
DFG (2014-2018, main applicant)
This project tries to bridge the gap between computational linguistics and theoretical linguistics by using linguistically-informed models and explicitly testing hypotheses stemming from Linguistics literature. It proposes a compositional approach to noun-noun (N-N) compound analysis with an interdependent three-level model that comprises compound splitting, capturing the meaning of the components and the covert relation that holds between them. We used multi-lingual data throughout the project, in analysis and evaluation, and followed a language-independent approach, using automatic, knowledge-lean, data-driven methods.
Collaborations with Gianina Iordachioaia (Institute for English Linguistics)
CLASSiC project: Cross-lingual semantic annotation transfer from English to French
EC FP7 (2008-2011, post-doc)
In the CLASSiC project (Computational Learning in Adaptive Systems for Spoken Conversation) we are focusing on semantic role labeling for French and in particular on methods to automatically generate semantic annotations for French. Syntactic annotation is available for French, but no semantic information. Since there is semantic annotation available for English and there are parallel corpora for the language pair English-French, we transfer the semantic annotation from English to French translations using word alignments. Contrary to previous work (Padó and Pitel, TALN 2007; Padó and Lapata, Comp. Ling. 2009; Basili et al. CICLing 2009), we did not use an ontology constructed for the target language. We want to minimize the amount of manual labour and aim for broad coverage annotations. We used the PropBank annotation framework constructed for English to annotate French sentences, after having tested the cross-lingual validity of PropBank (Van der Plas et al., LAW 2010). Because we know that there is a high correlation between syntax and semantics (see also Merlo and Van der Plas, ACL 2009), we leveraged the information contained in the syntactic annotations in a second step. In this step we trained a syntactic-semantic parser on the combination of syntactic annotations and the semantic annotations resulting from transfer. The automatically generated semantic annotations for French are close to the upper bound from manual annotations (Van der Plas et al., ACL 2011).
PhD project: Automatic lexico-semantic acquisition for question answering (NWO IMIX: 2003-2008, PhD student)
(Promotor: John Nerbonne, co-promotor: Gosse Bouma)
Freedom and liberty share the same meaning. Paris denotes a city, and the word party triggers associations of wine and fun for many. People naturally acquire these lexico-semantic relations such as synonyms, categorised named entities, and associations by using language in their daily life.
For many natural language processing applications, such as question answering, this type of information is essential, e.g. to recognise that a particular meaning can be inferred from different text variants or to compensate for the lack of general world knowledge.
This thesis proposes three methods for using large text corpora to acquire lexico-semantic information automatically: a syntax-based method, a multilingual word-alignment-based method and a proximity-based method. The three methods complement each other in the type of data needed, the way they deal with sparse data and most importantly, in the types of lexico-semantic information they provide. This information is then applied to the Groningen question answering system Joost. Among the different types of lexico-semantic information acquired, categorised named entities, e.g. Paris denotes a city, improved the system the most and this information was obtained with the syntax-based method.
In my spare time, I like to draw and paint: https://flic.kr/s/aHsj9J6viw, go hiking, swimming, and eat good food.