About ETRO  |  News  |  Events  |  Vacancies  |  Contact  
Home Research Education Industry Publications About ETRO

Master theses

Current and past ideas and concepts for Master Theses.

Neural architectures for joint entity and relation extraction


In this thesis, we focus on tasks from the NLP area.
NLP lies in the intersection of machine learning and computational linguistics (i.e., modeling of natural language).
NLP enables computers to process unstructured text and extract useful (e.g., structured) information.
To do so, in this master thesis, we will work in enriching sequences of text with indications of pre-defined entities (for example, names of people or organizations, or types of rooms in a house) and relations between these (e.g., works for –an example of the entity recognition and relation extraction problem is illustrated in Fig. 1).
These tasks, called named entity recognition and relation extraction, are core NLP tasks, and they have seen a lot of research in the past. However, a number of issues with existing methods leave several opportunities for fundamental further research steps.

Kind of work

In this thesis, we will work on several publicly available datasets for named entity recognition and relation extraction [Bekoulis, G., et al. (2018)] such as ACE04, DREC, ADE. Specifically, we will focus on (i) investigating existing baselines that consider the two tasks in a joint setting, (ii) identifying current limitations of state-of-the-art methods [Luan, Y., et al. (2019)], and (iii) alleviating those issues by developing new neural network architectures for entity and relation extraction.

Framework of the Thesis

Bekoulis, Giannis, et al. "Joint entity recognition and relation extraction as a multi-head selection problem." Expert Systems with Applications 114 (2018): 34-45.
Luan, Yi, et al. "A general framework for information extraction using dynamic span graphs." Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). 2019.

Number of Students


Expected Student Profile

Proven programming experience (e.g., Python)
Prior experience with state-of-the-art machine learning frameworks (e.g., Tensorflow, PyTorch)
Familiarity with the Natural Language Processing (NLP)


Prof. Dr. Ir. Nikolaos Deligiannis

+32 (0)2 629 1683

more info


Dr. Giannis Bekoulis

+32 (0)2 629 1686

more info


An example sentence for relation extraction where the two entities are coloured in green and blue, and the type of the relation is coloured in red.

- Contact person




- Contact person

- Thesis proposals

- ETRO Courses

- Contact person

- Spin-offs

- Know How

- Journals

- Conferences

- Books

- Vacancies

- News

- Events

- Press


ETRO Department

Tel: +32 2 629 29 30

©2022 • Vrije Universiteit Brussel • ETRO Dept. • Pleinlaan 2 • 1050 Brussels • Tel: +32 2 629 2930 (secretariat) • Fax: +32 2 629 2883 • WebmasterDisclaimer