Research

Introduction

I am a PhD candidate working in the computational linguistics group at the University of Groningen. My research focuses on text generation from abstract dependency structures, under supervision of dr. Gertjan van Noord. This research is conducted as a part of the STEVIN DAISY project. Our task within the DAISY project is to build a sentence generator for Dutch.

Interests

Sentence realization

Over the years wide-coverage grammars have been developed for some languages. I am interested in using these grammars for sentence realization.

Fluency ranking

Sentence realizers can produce sentences that are grammatical, but not necessarily fluent. I have an interest in developing models for ranking sentences by fluency.

Feature selection

Feature-based models are often created using templates that create a huge number of features. Feature selection techniques can be applied to pick an effective subset of features.

Sentence compression

Sentences can be shortened for various tasks, such as summarization or headline generation. I am interested in sentence compression methods that are constrained by a grammar.

Study

Contact information

Daniël de Kok
Computational Linguistics
University of Groningen
Oude Kijk in ’t Jatstraat 26
9712 EK Groningen
The Netherlands

E-Mail: d.j.a.de.kok%rug.nl