Modelling and Measuring Information Density (B04)

• Subject Area: Applied Linguistics, Computational Linguistics ; General and Comparative Linguistics, Experimental Linguistics, Typology, Non-European Languages ; Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing
• Term: since 2014
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 232722074

The main goal of B4 in the third phase of the CRC is to study language models for dynamically changing tasks tightly relating to the theme of “information in flux”. We research new language modeling techniques that can cope with temporally changing or drifting data and evaluate them in terms of perplexity. In addition, we explore resource usage and ways to change the resource use in language models. In particular, we look into ways to reduce memory consumption in a language model without sacrificing perplexity.

• DFG Programme: Collaborative Research Centres
• Subproject of: SFB 1102: Information Density and Linguistic Encoding
• Applicant Institution: Universität des Saarlandes
• Project Head: Professor Dr. Dietrich Klakow