(Ref. BAP-2016-55)
FULL-TIME SENIOR RESEARCHER IN NATURAL LANGUAGE PROCESSING, KU LEUVEN (UNIVERSITY OF LEUVEN) & IMINDS
The research team itec is looking for a full-time senior researcher in natural language processing (NLP) to analyze and extract information from large collections of unstructured written data. The candidate will mainly take on responsibilities in the project entitled “Adaptive Semantics Induction for Multilingual matching in an enriched human resOurces enVironment” (ASIMOV), and is expected to interact with other projects in the research team that are situated in or verge on the domain of NLP. Depending on his/her qualification and experience, the candidate can progressively take on coordinating responsibilities for our NLP projects.
itec (Interactive Technologies) is a research group dedicated to the design, development, and evaluation of software for automated decision support, aimed at helping people to make better decisions on the basis of data analysis, (meta-)heuristics, and visualization. Researchers at itec conduct both fundamental and applied research, often in close collaboration with industrial partners in sectors such as ICT, media, education and training, health, logistics, and manufacturing. They are active in two areas: technology-enhanced learning and operational optimization.
itec is affiliated with KU Leuven (University of Leuven) and iMinds, Flanders’ strategic research centre for digital innovation. It has offices in Kortrijk (KU Leuven campus Kulak) and Ghent (KU Leuven Technologiecampus Gent).
iMinds is Flanders’ digital research & entrepreneurship hub. It combines the strength of its 900+ researchers at 5 Flemish universities to conduct strategic and applied research in areas such as ICT, Media and Health.
Together with its research partners – companies, governments and non-profit organizations – iMinds translates digital know-how into concrete products and services.
In addition, iMinds supports researchers, entrepreneurs and start-ups in the successful market introduction of their ideas.
Responsibilities
The ASIMOV project will explore methods for automating the development process of semantic engines combined with the leveraging of behavioral data to increase the performance of job matching software. The ambition is that these semantic engines can learn ontologies from data in multiple languages (Dutch, English, French, German) instead of using manually created and translated ontologies, and can be adapted with relative ease to new domains (e.g. the automotive industry). Further, the job matching engine should be able to take into account user profile information extracted from applicants’ clicking behavior on job sites.
ASIMOV is funded by iMinds-ICON and involves research groups from iMinds active in the field of data science (UGent – Data Science Lab and itec), as well as Flemish companies active in the domains of job matching, machine translation, and publishing. ICON projects involve partners from industry and research in a strong interdisciplinary and collaborative format, with a view to developing demonstrators that are expected to generate societal and economic impact after the project. ASIMOV starts on 1 April 2016 and ends 31 March 2018.
The responsibility of the candidate in this project is to lead and carry out a work package that concentrates on inductive methods (automation) for extracting and adapting multilingual ontologies. The candidate is also expected to contribute to otherwork packages. Tasks involve:
- Investigating,implementing and evaluating methods for learning jobs and skills from data using part-of-speech tagging and (shallow) parsing, including methods for detecting multi-word units and for treating these as tokens;
- Investigating,implementing and evaluating methods for mapping (multi-word) terms between different languages;
- Publishing about this research in academic venues;
- Contributing to the implementation of the developed methods in a demonstrator;
- Contributing to project management, communication, and reporting.
Profile
- The applicant must hold a PhD or an MA degree (with working experience) in NLP.
- Solid experience in the (semi-)automatic cleaning and alignment of (parallel)corpora.
- Excellent knowledge of part-of-speech tagging and (shallow) parsing, including software libraries such as Apache OpenNLP, DkPro-Core, GATE and NLTK.
- Good skills in programming languages (such as Java or Python) and scripting languages (such as Perl or Python), and/or willingness to (further) develop these skills.
- Excellent knowledge of Unix/Linux platform and basic text processing tools (sed, grep,awk, etc.).
- Experience in server-side web application development and databases is an asset.
- Excellent command of English. Knowledge of Dutch is considered an asset.
Offer
- Length:1 year, renewable for 1 year.
- Seniority:depending on the degree and professional experience of the applicant.
- Beginningof the contract: from 1 April 2016.
- Work place: KU Leuven campus Kulak Kortrijk.