Seminar für Sprachwissenschaft

Forschungsinteressen

My research seeks answers to the question of how we, as humans, understand speech. I am, more specifically, interested in studying the processing of speech that entails decoding the speech signal and mapping the input acoustic information to an abstract representation of meaning. I approach this question through the discriminative learning framework. My work is currently focused on the development of naive and linear descriminative computational models of speech perception.

Lehre

SS 2018
Mathematical Methods: Statistics, Undergraduate course, Eberhard Karls University, Tübingen, Germany (2018)

Mathematical methods are essential for understanding and working in theoretical and computational linguistics. This course introduces the key concepts from the areas of set theory, algebra and logic, which belong to the basic repertoire of linguistic methods. The main goal of the course is to provide the students with sufficient competence in basic notations, terminology and concepts of discrete mathematics for their studies in theoretical and computational linguistics. Familiarity with concepts such as sets, functions and propositions, and the ability to work with simple proof techniques are a crucial prerequisite for subsequent courses. Students should acquire sufficient competence in basic notations, terminology and concepts of mathematics for their studies in linguistics. The topics of the course comprise the most essential mathematical notions needed in general linguistics, computational linguistics, document processing and information management. Familiarity with concepts such as sets, functions and propositions, and the ability to work with simple proof techniques will be expected in subsequent courses. The main purpose of the course is to equip the participants with the most basic mathematical tools which they will need in their linguistics courses. Verwendung: BA Allg. Sprachwissenschaft: Modul Methods II BA Computerlinguistik / Computational Linguistics: Modul Methods II

Veröffentlichungen

Shafaei-Bajestan, E. and Baayen, R. H. (2018). Wide learning for auditory comprehension. In Proceedings of Interspeech 2018. pdf

Bajestan, E. S., Frassinelli, D., Lapesa, G., & Padó, S. (2017). DErivCelex: Development and Evaluation of a German Derivational Morphology Lexicon based on CELEX. on Resources and Tools for Derivational Morphology (DeriMo), 117. pdf

Submitted / Accepted:

Baayen, R. H., Chuang, Y.-Y., Shafaei-Bajestan, E., and Blevins, J. (2018). The discriminative lexicon: A unified computational model for the lexicon and lexical processing in comprehension and production grounded not in (de)composition but in linear discriminative learning. Complexity. pdf

Denistia, K., Shafaei-Bajestan, E., Baayen, R. H. (2018). Semantic Vector Model on the Indonesian Prefixes pe- and peN-. In Proceedings of the Mental Lexicon.

Präsentationen

Baayen, R. H., and E. Shafaei-Bajestan, A discriminative perspective on lexical access in auditory comprehension, Basque Center for Applied Mathematics, Bilbao, Spain, April 10, 2018.

Baayen, R. H., and E. Shafaei-Bajestan, A discriminative perspective on lexical access in auditory comprehension and speech production, Basque Center on Cognition, Brain and Language, San Sebastian, Spain, April 9, 2018.