Researcher in Explainable AI and Medical Natural Language Processing

(ref. BAP-2022-571)

Laatst aangepast : 20/07/2022

We offer a two-year research position for a postdoctoral or predoctoral scientist on the topic of explainable AI in the context of natural language processing of clinical reports at the Department of Computer Science, KU Leuven. The position is financed by the CHIST-ERA project ANTIDOTE (ArgumeNtaTIon-Driven explainable artificial intelligence fOr digiTal mEdicine). The project focuses on deep learning models for text classification in the medical domain, where the need for high quality explanations of clinical decisions is critical. The project proposes an integrated approach for the understanding of textual medical data and the generation of a textual explanation to justify a medical diagnosis made by a neural network, so that clinicians can judge its validity. The partners of the ANTIDOTE project provide annotated argumentative structures of natural language explanations of medical decisions (human explanations). We will also leverage larger-scale data (without annotated explanations) in order to improve explanation generation. To guide explanation generation we will use feature importance techniques (integrated gradients, attention mechanisms, or search-based methods). The researcher will perform his or her research at the Language Intelligence and Information Retrieval lab (https://liir.cs.kuleuven.be), which is directed by Prof. Marie-Francine Moens and which is a member of the KU Leuven AI Institute (https://ai.kuleuven.be) and the KU Leuven ELLIS Unit. The European Lab for Learning and Intelligent Systems (ELLIS) (https://ellis.eu) promotes basic research in machine learning and modern AI. KU Leuven is declared the most innovative university in Europe for the fourth year in a row (Reuters) and is ranked 42th on the Times Higher Education World University Ranking (2022).

Responsibilities

• Perform fundamental research in machine learning (more specifically interpretable neural networks), natural language processing, and explainable AI resulting in publications in highly-ranked venues.

• Collaborate with the other European partners of the ANTIDOTE project.

Profile

• You have (or are near completion of) a master or PhD in Computer Science.

• Given that we have to fill in the research position very soon, we will only consider applicants who can work in the European Union without a visa application.

• You have a demonstrated expertise in machine learning, deep neural networks and natural language processing. Expertise in structured prediction and/or structured generation is a plus.

• You have an outstanding track record of publications in relevant international peer-reviewed A ranked conferences and in journals with high impact factor in the fields of machine learning and natural language processing.

• You are good at collaborating with others.

• You work proactively and independently and have good communication skills.

• You have an excellent knowledge of the English language, both spoken and written.

Offer

• We offer a research position for 2 years starting September 1 or the latest October 1, 2022.

• We offer a competitive wage and yearly budget to attend conferences.

• We offer the opportunity for personal development beyond research including the supervision of master and PhD students and the possibility to contribute to teaching.

• We offer access to our deep learning cluster which contains the latest GPU hardware.

Interested?

For more information please contact Prof. dr. Marie-Francine Moens, tel.: +32 16 32 53 83, mail: [email protected] or Mr. Damien Sileo, tel.: +32 16 19 39 61, mail: [email protected].

KU Leuven seeks to foster an environment where all talents can flourish, regardless of gender, age, cultural background, nationality or impairments. If you have any questions relating to accessibility or support, please contact us at [email protected].