Ongoing Research Projects

AcListant (Active Listening Assistant): LSV (spoken language systems) together with DLR Braunschweig (German Aerospace Centre) work on a project with the focus on controller-pilot communication in order to perform research on methods for integrating the contextual knowledge from the environment into the speech recognizer.
ALLEGRO: A collaboration between Saarland University, DFKI, LORIA in Nancy and Supelec in Metz, ALLEGRO is an applied natural language processing project whose goal is to investigate the use of NLP in building interactive dialogue-based foreign language learning exercises as well as tools which will allow content developers to create linguistic resources required for the exercises, at ease.
CoDIS: Compositionality and grounding for distributional semantics: statistical modeling of lexical meaning within phrases and sentences. CoDIS is a project in the excellence cluster Multimodal Computing and Interaction, funded by DFG (German Science Foundation).
Exemplar-based speech representation: Exemplar Theory: Exemplar-based speech representation (Möbius with Schütze, Univ. Stuttgart; funded by DFG within SFB 732 [A2])
Function of Speech Rhythm in Information Processing: Andreeva and Marti (funded by Saarland University)
Functional interpretation of healthy speech production: Analysis of cerebral organisational structures, and of acoustical and electrophysiological correlates of speech production. (Pützer, in cooperation with Saarland University Hospital, funded by Saarland University)
Individualized feedback in computer-assisted spoken language learning (IFCASL): Möbius and Trouvain (funded by DFG) and Laprie (funded by ANR)
LanPercept (Language and Perception): The Psycholinguistics Group is part of the Marie Curie Initial Training Network LanPercept (Language and Perception) to study language-vision mapping. The network will train 11 PhD researchers and 4 postdoctoral researchers across 8 leading academic and 7 industrial and private sector partners in Europe. The project starts in January 2013 and lasts for four years. (Crocker EU-funded)
METALOGUE: The goal of the project is to produce a multimodal dialogue system that is able to implement an interactive behaviour that seems natural to users and is flexible enough to exploit the full potential of multimodal interaction. (Klakow EU-funded)
Neurology of Speech Production: fMRI investigation of the neuroanatomical correlates of speech production in collaboration with the University Hospital, Homburg.
a) Initiation and execution of articulatory gestures in healthy speakers.
b) Effects of stimulation in sub-cortical structures on phonation and articulation.
(Pützer, internal funding)
PENGUIN: The project on planning-based efficient generation of instructions (PENGUIN) focuses on the real-time generation of natural-language instructions that help a user solve a task in a virtual environment. This research brings together work on natural-language semantics, discourse, and pragmatics in situated communication with work on AI planning and efficient algorithms. The project is funded by the Cluster of Excellence on Multimodal Computing and Interaction.
PHEME: Social media poses three major computational challenges, dubbed by Gartner the 3Vs of big data: volume, velocity, and variety. Content analytics methods have faced additional difficulties, arising from the short, noisy,and strongly contextualised nature of social media. In order to address the 3Vs of social media, new language technologies have emerged, e.g. using locality sensitive hashing to detect breaking stories from media streams (volume), predicting stock market movements from microblog sentiment (velocity and recommending blogs and news articles based on user content (variety). PHEME will focus on a fourth crucial, but hitherto largely unstudied, challenge: veracity. It will model, identify, and verify phemes (internet memes with added truthfulness or deception), as they spread across media, languages, and social networks. (Uszkoreit EU-funded)
PIRE: The partnership for research and education (PIRE), established in 2005, is a collaborative PhD programme between Saarland University, Germany, Charles University, Prague, the Brown Laboratory for Linguistic Information Processing headed by Prof. Eugene Charniak and The Johns Hopkins University Center for Language and Speech Processing (CLSP) headed by Prof. Frederick Jelinek.
ROCKIT: ROCKIT is a strategic roadmapping project for research and innovation in the area of natural conversational interaction. The primary scientific focus concerns interactive agents which are proactive, multimodal, social, and autonomous. A second focus concerns systems which can extract and exploit rich context and knowledge from heterogeneous data sources (Klakow EU-funded).
SCALE: The SCALE project is concerned with adaptive learning approaches to all areas of speech processing, with particular focuses on automatic speech recognition and synthesis, signal processing, human speech recognition, and machine learning.
Schreibgebrauch: Manfred Pinkal's group will collaborate with the Institute for German Language (IdS Mannheim), the editorial offices of Duden and Wahrig, and the Deutsche Rechtschreibrat from 2013 to 2016 on a project funded by BMBF (the German Federal Ministry of Education and Research The project will provide an infrastructure for the corpus-based investigation of ongoing changes in German spelling, to provide an empirical basis for future adaptations of official spelling rules. Saarland University will work on the development and adaptation of language technology tools, in particular for application to non-standard language, as it can be found in genres like email and Twitter.
SMILE: The SMILE project - Script Mining as Internet-based Learning - centers around learning typical structures & participants of common-sense activities. SMILE is funded by the Cluster of Excellence "Multimodal Computing and Interaction".
The syllable as a processing unit: Speech Production: The syllable as a processing unit (Möbius with Wagner, Univ. Bielefeld; funded by DFG within SPP-1234)
Tonal and temporal development of L1 prosodic categories: Prosody Acquisition: Tonal and temporal development of L1 prosodic categories (Möbius; funded by DFG)

Completed Research Projects

audite vocem: Online training program for perceptual identification of different voice qualities.
(Köser, Pützer; funded by Saarland University)
BiLD: Bidirectional Linguistic Deduction
Bulgarian online language learning: Bulgarisch: Entwicklung und Durchführung von multimedialen Fremdsprachenkursen (eLearning)
(Andreeva, Avgustinova; funded by Saarland University)
CHORUS: Semantic Processing With Constraints and Preferences
COLLATE: Computational Linguistics and Language Technology for Real Life Applications
CORTE: Computerlinguistische Methoden für die Rechtsterminologie
DANDELION: Discourse Functions and Discourse Representation
Deep Thought: Hybrid and Shallow Methods for Knowledge-Intensive Information Extraction
DIALOG: Tutorial Dialogue with a Mathematics Assistance System
ELAN: Ellipsis and Anaphor
ESPERONTO: Bridging the gap between WWW and Semantic Web
eTon: The music of speech: Multimodal training in production and perception of pitch contours in Standard German with regard to various functions in communication. (Andreeva, Köser, Gast; funded by Saarland University)
FraCaS: Framework for Computational Semantics
GENIE: Datenbank Gesprochenes Niedersorbisch/-wendisch, in collaboration with the slavonics department.; (Marti, von Rauch, Barry, Andreeva; funded by Saarland University, 2005-2006)
InDiGen: Integrated Discourse Generation
Intera: Building an integrated European language resource area by connecting international, national and regional data centres, and producing new multilingual language resources.
IRTG: The International Research Training Group (formerly International Post-Graduate College) "Language Technology and Cognitive Systems" offered a joint post-graduate education programme of Saarland University and the University of Edinburgh.
LATER: Language Technology Erasmus Mundus Programme
LATESLAV: Language Processing Technologies for Slavic Languages
LILOG: Linguistic and Logical Methods for Text Understanding
LISA: Inference and Natural Language Semantics
MiLCA: Medienintensive Lehrmodule in der Computerlinguistik-Ausbildung
MULI: Annotation of Information Structure in German and English text corpora
NEGRA: Concurrent Grammatical Processing
PERFORM: Performance Modelling for Declarative Grammars
Processing Partially Free Word Order
Prominence & Rhythm: Cross-language analysis of the production and perception of prominence.
(Barry, Andreeva, Kötzer; funded by DFG, withinSPP1234)
PTOLEMAIOS: Emmy Noether Research Group on Grammar Learning from Parallel Corpora
RGR: Reusable Grammatical Resources
SALSA: The Saarbrücken Lexical Semantics Annotation Project: Creating a corpus annotated with semantic roles, and utilizing it for the construction of a big semantic lexicon.
SAMOS: Semantic Analysis in a Modular System
SFB 378: Collaborative Research Centre 378: Resource-Adaptive Cognitive Processes
SIRIDUS: Specification, Interaction and Reconfiguration in Dialogue Understanding Systems
TALK: Development of adaptive multimodal dialogue systems using the "Information State Update" (ISU) approach to dialogue management, originating from the TRINDI and SIRIDUS projects.
TIGER: Linguistic Interpretation of a German Corpus
TRINDI: Task Oriented Instructional Dialogue
Verbmobil