I am a Machine Learning Architect at LynxCare, where I build models for clinical natural language processing. Previously, I was a research scientist at Nuance Automotive (now Cerence) and a postdoctoral researcher at IDLab's Text-to-Knowledge Group at Ghent University. My research focused on training and deploying neural NLP models in low-resource settings, with work spanning sequence-to-sequence models, keyphrase extraction, and knowledge base population.

Feel free to reach out by mail or connect on social media.

Publications

Learning to Reuse Distractors to Support Multiple Choice Question Generation in Education
S. K. Bitew, A. Hadifar, L. Sterckx, J. Deleu, C. Develder, T. Demeester
IEEE Transactions on Learning Technologies
Clinical Information Extraction for Preterm Birth Risk Prediction
L. Sterckx, G. Vandewiele, I. Dehaene, O. Janssens, F. Ongenae, F. De Backere, F. De Turck, K. Roelens, J. Decruyenaere, S. Van Hoecke, T. Demeester
Journal of Biomedical Informatics
Overly Optimistic Prediction Results on Imbalanced Data: Flaws and Benefits of Applying Over-sampling
G. Vandewiele, I. Dehaene, G. Kovacs, L. Sterckx, O. Janssens, F. Ongenae, F. De Backere, F. De Turck, K. Roelens, J. Decruyenaere, S. Van Hoecke, T. Demeester
Artificial Intelligence in Medicine, 2021
A Self-Training Approach for Short Text Clustering
A. Hadifar, L. Sterckx, T. Demeester, C. Develder
RepL4NLP at ACL 2019
Predicting Psychological Health from Childhood Essays. The UGent-IDLab CLPsych 2018 Shared Task System
K. Zaporojets, L. Sterckx, J. Deleu, T. Demeester, C. Develder
CLPsych at NAACL-HLT 2018
Prior Attention for Style-aware Sequence-to-Sequence Models
L. Sterckx, J. Deleu, C. Develder, T. Demeester
arXiv preprint
Break it Down for Me: A Study in Automated Lyric Annotation
L. Sterckx, J. Naradowsky, B. Byrne, T. Demeester, C. Develder
EMNLP 2017
Creation and Evaluation of Large Keyphrase Extraction Collections with Multiple Opinions
L. Sterckx, T. Demeester, J. Deleu, C. Develder
Language Resources and Evaluation
Supervised Keyphrase Extraction as Positive Unlabeled Learning
L. Sterckx, C. Caragea, T. Demeester, C. Develder
EMNLP 2016
Knowledge Base Population using Semantic Label Propagation
L. Sterckx, T. Demeester, J. Deleu, C. Develder
Knowledge-Based Systems
An Automated End-To-End Pipeline for Fine-Grained Video Annotation using Deep Neural Networks
B. Vandersmissen, L. Sterckx, T. Demeester, A. Jalalvand, W. De Neve, R. Van de Walle
ICMR 2016 (Demo)
Ghent University-IBCN Participation in TAC-KBP 2015 Cold Start Slot Filling Task
L. Sterckx, T. Demeester, J. Deleu, C. Develder
TAC 2015
Topical Word Importance for Fast Keyphrase Extraction
L. Sterckx, T. Demeester, J. Deleu, C. Develder
WWW 2015 (Poster Session)
When Topic Models Disagree: Keyphrase Extraction with Multiple Topic Models
L. Sterckx, T. Demeester, J. Deleu, C. Develder
WWW 2015 (Poster Session)
Using Active Learning and Semantic Clustering for Noise Reduction in Distant Supervision
L. Sterckx, T. Demeester, J. Deleu, C. Develder
AKBC Workshop at NIPS 2014
Ghent University-IBCN Participation in TAC-KBP 2014 Slot Filling and Coldstart Tasks
M. Feys, L. Sterckx, L. Mertens, J. Deleu, T. Demeester, C. Develder
TAC 2014
Assessing Quality of Unsupervised Topics in Song Lyrics
L. Sterckx, T. Demeester, J. Deleu, L. Mertens, C. Develder
ECIR 2014

Thesis

Methods for Efficient Supervision in Natural Language Processing
L. Sterckx

Reports

An Evaluation of Neural Network Models for Toxic Comment Classification
L. Sterckx

Talks

Information Extraction from Medical Notes for Early Birth Risk Prediction
12th Belgium NLP Meetup — December 2019
Knowledge Base Population from Text and Graphs
Cambridge Language Technology Lab, Seminar — February 2017
Distributed Representations of Relation Paths to Bootstrap Relation Extractors
UCL Machine Reading Lab, Internal Workshop — July 2016