Absicht

BulDialects is a joint project of the Alfa-Informatica department of RU
Groningen (Netherlands) and the SfS (Seminar für Sprachwissenschaft) at the
University of Tübingen, with the assistance of the Bulgarian Academy of
Sciences.

This project will use both established and novel methods to perform
quantitative analysis of the dialects of Bulgarian and of adjacent territories.
Later in the project, these methods will be extended to other European
languages.

Project aims

BW-eSci(T) is a joint project of the University of Tübingen and the Fachinformationszentrums (FIZ) Karlsruhe. Within the university, the Department of General and Computational Linguistics (SfS), as well as the Center for Information, Communication and Media (IKM) are involved.

The goal of the project is to develop a prototypical eScience research environment for the University of Tübingen. For researchers it is especially important that all data is preserved an made available on a long-term basis. The eSciDocsoftware can be used to meet these requirements.

eSciDoc is an infrastructure platform for storing and maintaining of primary research data as well as analyzations of data and publications. It was primarily developed to provide researchers with open access to data, tools, and other resources. Furthermore, this project will ease interdisciplinary scientific research.

Abstract

CLAIRE (Constraint Linguistics: An Interdisciplinary Review and Evaluation) is a collaborative project between the Seminar für Sprachwissenschaft of the University of Tübingen and the Computational Linguistics/Machine Translation Group (CL/MT Group) of the University of Essex, UK, funded by the Deutscher Akademischer Austauschdienst(DAAD) and the British Council, with additional support from the Anglo-German Foundation for the Study of Industrial Society, in the framework of the Academic Research Collaboration Programme (ARC), which aims at facilitating collaborative links between research groups at publicly funded institutions in the United Kingdom and Germany.

Project aims

The CLARA (Common Language Resources and their Applications) project is a Marie Curie Initial Training Network in the area of Common Language Resources and their Applications.

It provides early stage and experienced researchers with scholarships for specific research tasks and for specific durations. To connect European research and to exchange knowledge between research institutions, researchers are exchanged between European countries.

CLARA will train a new generation of researchers who will be able to cooperate across national boundaries on the establishment of a common language resources infrastructure and its exploitation for the construction of the next generation of language models with wide theoretical and applied significance. The scientific objectives of the CLARA research context are twofold:

• to develop the next generation of data-intensive language models and applications by integrating approaches across language and country boundaries; 
• to contribute to the establishment of a pan-European infrastructure for language resources. 

CLARA will supplement basic research competencies in the language and text sciences with specialized knowledge and skills in linguistics, computer science, knowledge engineering, databases, statistical processing and language and speech applications. Participation in advanced research at leading universities will be complemented with additional training at industrial partners contributing to careers in industry.

Absicht

CLARIN-PLUS was dedicated to enhancing CLARIN. Following the recommendations of the 2013 ESFRI
Assessment Expert Group, CLARIN-PLUS aimed to accelerate the implementation and to strengthen and
consolidate CLARIN in the following areas:

1. The central (technical) hub
2. The central office
3. Partnerships with other infrastructures
4. Outreach
5. Governance

The EKUT part of the CLARIN-PLUS project centered around the CLARIN Language Resource Switchboard
(see https://switchboard.clarin.eu). This browser-based software enables users to easily identify and start
NLP applications that can process their language ressources in one way or another.

The project ran from September 2015 to Dezember 2017.

Abstract

The Tübingen-Sofia International Graduate Programme in Computational Linguistics
and Represented Knowledge (CLaRK) provides a joint teaching and research
facility for doctoral and master's students from Central and Eastern Europe.
CLaRK is funded by the Volkswagen Foundation.

Abstract

Adapting Bilingual Dictionaries for On-Line Comprehension Assistance

Abstract

The D-SPIN Project (Deutsche Sprachressourcen-Infrastruktur), funded by the
German Federal Ministry of Education and Research, is the German contribution
to the European CLARIN Project.  The CLARIN project is oriented towards the
language resource needs of humanities researchers and places an emphasis on the
training of young researchers.  Additionally, the project collects language
resources in Germany.  This project is coordinated by the SfS (Seminar für
Sprachwissenschaft) at the University of Tübingen, which contributes to nearly
all project workpackages.

Abstract

The goal of the DEREKO project is to assemble a comprehensive German written
language reference corpus, including annotation, and to develop appropriate text
analysis tools.  It is funded by the Baden-Württemberg Ministry of Science,
Research and the Arts.

Abstract

Dynamic interpretation of natural language

Project aims

 EUDAT is a three-year project that will deliver a Collaborative Data Infrastructure (CDI) with the capacity and capability for meeting future researchers’ needs in a sustainable way. Its design will reflect a comprehensive picture of the data service requirements of the research communities in Europe and beyond. 

 Although the user requirements will vary between scientific disciplines, the micro-systems from which each community’s services are built are largely generic. With the inclusion of disciplines from across the spectrum of scientific endeavor sharing a common infrastructure, EUDAT also provides the opportunity for data-sharing between disciplines and cross-fertilization of ideas. 

 As part of the CLARIN user community, the Tübingen research group is participating as a contributor of linguistic use-cases for the CDI. 

Abstract

Corpus-based development of lexical knowledge bases.

Abstract

Programme to fund doctoral students from Eastern and Central Europe..

Abstract

The goal of the EuroWordNet project is the creation of a multilingual
lexico-semantic database of words and their meanings.  In this database, the
meanings of words are digitally encoded and linked together through various
relations including synonymy, antonymy, and hyponymy.  These concepts and
their relations form a semantic network.  The multilingual database furthermore
connects concepts and lexical items across the individual language projects
(for German, the GermaNet network).  That way, EuroWordNet can be used as a
resource for multilingual information retrieval applications.

The first stage of the project will integrate English, Dutch, Italian and
Spanish into the database.  The second phase (ending June 1998) will extend the
database to German, French, Estonian and Czech.

Abstract

Construction of the GermaNet lexical-semantic network for German, funded by the
Baden-Württemberg Ministry of Science, Research and the Arts.

Absicht

German-Israeli Minerva School in Computational Linguistics in Tuebingen and Blaubeuren.

Project aim

The Competence Center Computational Linguistics is a joint project of
the Institute for Natural Language Processing at the University of
Stuttgart and the SfS (Seminar für Sprachwissenschaft) of the University
of Tübingen.  This project, running from early 2000 til 2005, was funded
by the Baden-Württemberg Ministry of Science, Research and the Arts and
the universities of Stuttgart and Tübingen.

With the establishment of the Competence Center, we intend to raise
awareness of the industrial applications of computational linguistics.
We hope to actively contribute to the transfer of knowledge from
academic to industrial contexts, beginning with the resources and
procedures developed at the two sponsoring institutes.

Project aims:

KobRA will use innovative data mining methods to advance structured corpus-based research in empirical linguistics (research using annotated text corpora, treebanks, digital lexical, etc.) This project will apply machine learning to existing resource infrastructure in order to improve, structure, and visualize search results for users, and to develop new computational techniques for the social sciences.  We will establish a high level of cooperation between computer scientists, language technologists and linguists for the purpose of uncovering new insights into the relationship between linguistic annotations and machine learning techniques, and especially discovering what kinds of annotations and corpus structures are best to suited to particular computational research techniques.

Abstract

STEEL (1996-1999) was an EU Copernicus Programme project that developed tools
for translating Eastern European languages.

Abstract

Given the enormous amount of static and dynamically generated material
developed for eLearning applications, the biggest challenge to enhancing
the usability of these resources is improving access and information
retrieval in learning management systems.

The goal of the LT4eL project (Language Technology for eLearning) is to
solve this problem by using language technologies and integrating semantic
knowledge to improve the management, dissemination, and searchability of
learning materials.

The project will specifically devise resources and tools to semi-automatically
generate descriptive metadata. It will develop new technologies like keyword
extractors and glossary candidate detectors for project's target languages:
Bulgarian, Czech, Dutch, English, German, Maltese, Polish, Portuguese and

Abstract

LTfLL is a EU-funded eLearning project aiming to bring together language
technology, computer-mediated learning and self-directed education to
enable users to find their own optimal way to learn, based on their own
background and abilities.  The SfS (Seminar für Sprachwissenschaft) is
responsible for the workpackage concerned with user orientation at the
beginning and later stages of the learning process, using LSA technologies
to evaluate user progress.

Abstract

MiLCA (Medienintensive Lehrmodule in der Computerlinguistik-Ausbildung) is a
pilot project for training in computational linguistics, funded by the German
Federal Ministry of Education and Research.

Project aims

Linguistic resources such as corpora, lexicons, grammars, experimental data or computer programs are important in linguistic research. The creation of such data is often a very complicated process, and in the long run data can be lost or can no longer be worked with. The NaLiDa project deals with developing an infrastructure for long-term sustainability of linguistic data, so that such data will be accessible for conducting research in the future.

Abstract

In the framework of the PROCOPE programme, which sponsors cooperation between
French and German researchers, the DAAD (German Academic Exchange Service)
supports cooperation between computational linguists at the University of
Tübingen and the Laboratoire Langue, Langages et Cognition (2LC) at the
University of Aix-en-Provence.

Abstract

A basis in linguistic theory and computational implementation of
domain specific processing for constraint-based grammars.

Project aims

This project is concerned with the annotation of complex sentence
structures, focusing on grammatical functions like parataxis, hypotaxis,
and coordination, as well as anaphora resolution.

Project aims

The C1 Project addresses two of the central concerns of the SFB:
technology transfer within the SFB, and the link between
methodological/theoretical disciplines and phenomenological approaches,
particularly with respect to research results and technology
development.  C1 is furthermore responsible for technology transfer to
the general public.  C1 will ensure that SFB-developed data and
annotation resources will be accessible and reusable for scientists
outside of the SFB via the Internet as part of the planned TUSNELDA
(Tübinger Sammlung nutzbarer empirischer linguistischer Datenstrukturen)
project.

Project aims

SFB joint projects 441, 538 and 632 created digital linguistic data
collections for research and general linguistics use.  These empirical
data sources are of great benefit to linguistic and philological
research.  The C2 project has the goal of creating conditions that will
promote the permanent and sustainable availability of these data sources
beyond the end of SFB project support.

Project aims

The two most important goals of the “Semantic Information Retrieval” project are

• to build a large-scale lexical-semantic translation resource that contains sense disambiguated data from Wikipedia, Wiktionary, and EuroWordNet, and 
• to implement new lexical-semantic methods for cross-lingual information retrieval. 

Disambiguating word senses of search terms is a special challenge, because polysemous terms may result in incorrect translations and hence to irrelevant documents for a search. The extension of GermaNet and its linking to EuroWordNet are key aspects for the creation of this translation resource. With the help of this resource, new methods for cross-lingual information retrieval will be developed and analyzed, which are based on trained statistical translation models and which aim at improving automatic translation systems.

Abstract

The Semantic Information Retrieval Project, a joint project of the SfS (Seminar
für Sprachwissenschaft) at the University of Tübingen and the eLearning
research unit of TU-Darmstadt funded by the DFG (German Research Foundation),
aims to improve information retrieval for computer-assisted career counseling.
It will employ semantic similarity measures determined by the semantic and
lexical relations between words in queries and stored documents.  New relations
between domain-specific concepts, discovered through unsupervised machine
learning from annotated corpora of job descriptions, will be documented in an
expanded version of the GermaNet lexico-semantic framework.

Abstract

Lexical-Semantic Net for German

Abstract

The project Syntax-based Coreference and Anaphora Resolution (SynCAR) is an exchange program between the Research Group in Computational Linguistics at Wolverhampton, UK, and the Seminar für Sprachwissenschaft at Tübingen. It concentrates on research concerning synrtax-based methods for anaphora and coreference resolution.

Absicht

Learning Computational Grammars (LCG) is a project funded by the European Community's Training and Mobility of Researchers programme. The goals of the network are to explore machine learning techniques for natural language. There will be a focus on comparison of results based on attempts to learn noun phrase structure in English.

Abstract

Joint project of the German Federal Ministry of Education and Research into
automatic translation for spoken language.

Absicht

ViCE – the Virtual Open Science Collaboration Environment was a
cooperation project of academic institutions in the state of
Baden-Württemberg, coordinated by the computing centre of the University of Freiburg. In this project, the requirements of disciplinary researchers and the services available at computing centres were combined, enabling computing centers to offer new services which apply advanced technologies to applications used in disciplinary research. In this context, the General and Computational Linguistics Department at the University of Tübingen supported the English Department at the University of Freiburg in applying technologies offered by the computing
centres to applications used in the humanities.

The goal of the project ViCE was, through close cooperation between the community and the computing centres, to put English Studies researchers in a position to make use of commercially available applications (e.g. Google, Chadwyck-Healey, etc.), as well as services offered by national and European infrastructures (such as CLARIN and DARIAH), together with services and technologies offered by the computing centers. Since all of
these services are used by English Studies researchers, it was necessary to combine them in order to meet the technical requirements of the community.

A major result of the project was the technical integration of the
portal "Baden-Württemberg Digital English Studies Community" (BW-DESC, https://bw-desc.de/) into the virtual environment of the technology partners. Additionally, through this project, all students, scholars and teachers of the English Studies programs in the state of Baden-Württemberg now have online access to large data collections, stemming from both European research infrastructures and partially commercial collections such as corpus.byu.edu.

Abstract

VirtuGrade is a project to research and implement digital
media-supported education and graduate teaching.  The project will
support the "Virtuelle
Hochschule" (Virtual University) programme launched by the Baden-Württemberg
Ministry of Science, Research and the Arts to promote the use of new media in
education.