UCT NLP

University of Cape Town Natural Language Processing Group

We perform research in natural language processing and machine learning. Research topics include:

• Methods for text generation and language model training for low-resource languages
• Approaches for modelling linguistic structure in text
• Creating NLP datasets and models for South African languages

Principle Investigator

Dr. Jan Buys

Current postgraduate students

• Francois Meyer (PhD)
• Sello Ralethe (PhD)
• Nomonde Khalo (PhD)
• Claytone Sikasote (PhD)
• Natalie Bianca Alexander (MSc)
• Khalid N. Elmadani (MSc)

Past students

• Neil Sinclair (MSc)
• Shane Acton (MSc)
• Victoria Pedlar (MSc)
• Yassin Nurmahomed (MSc)
• Maxwell Mojapelo (MPhil)

Information for Prospective Students

Publications

• A Systematic Analysis of Subwords and Cross-Lingual Transfer in Multilingual Translation.
  Francois Meyer and Jan Buys.
  NAACL Findings 2024.
  
  
• NGLUEni: Benchmarking and Adapting Pretrained Language Models for Nguni Languages.
  Francois Meyer, Haiyue Song, Abhisek Chakrabarty, Jan Buys, Raj Dabre and Hideki Tanaka.
  LREC-COLING 2024. [Data]
  
  
• Triples-to-isiXhosa (T2X): Addressing the Challenges of Low-Resource Agglutinative Data-to-Text Generation.
  Francois Meyer and Jan Buys.
  LREC-COLING 2024. [Data]
  
  
• Neural Machine Translation between Low-Resource Languages with Synthetic Pivoting.
  Khalid N. Elmadani and Jan Buys.
  LREC-COLING 2024.
  
  
• Subword Segmental Machine Translation: Unifying Segmentation and Target Sentence Generation.
  Francois Meyer and Jan Buys.
  ACL Findings 2023. [Code]
  
  
• Data Augmentation for Low Resource Neural Machine Translation for Sotho-Tswana Languages.
  Maxwell Mojapelo and Jan Buys.
  SACAIR 2023.
  
  
• Policy-based Reinforcement Learning for Generalisation in Interactive Text-based Environments.
  Edan Toledo, Jan Buys and Jonathan Shock.
  EACL 2023. [Code]
  
  
• Subword Segmental Language Modelling for Nguni Languages.
  Francois Meyer and Jan Buys.
  EMNLP Findings 2022. [Code]
  
  
• University of Cape Town’s WMT22 System: Multilingual Machine Translation for Southern African Languages.
  Khalid N. Elmadani, Francois Meyer and Jan Buys.
  WMT 2022. [Model]
  
  
• Self-Supervised Text Style Transfer with Rationale Prediction and Pretrained Transformers.
  Neil Sinclair and Jan Buys.
  SACAIR 2022 (CCIS). [Version of Record]
  
  
• From GNNs to Sparse Transformers: Graph-based architectures for Multi-hop Question Answering.
  Shane Acton and Jan Buys.
  SACAIR 2022 (CCIS). [Version of Record]
  
  
• Generic Overgeneralization in Pre-trained Language Models.
  Sello Ralethe and Jan Buys.
  COLING 2022. [Data]
  
  
• A Sequence Modelling Approach to Question Answering in Text-Based Games.
  Greg Furman, Edan Toledo, Jonathan Shock and Jan Buys.
  Wordplay 2022. [Code]
  
  
• Canonical and Surface Morphological Segmentation for Nguni Languages.
  Tumi Moeng, Sheldon Reay, Aaron Daniels and Jan Buys.
  SACAIR 2021 (CCIS). [Version of Record] [Code]
  
  
• Low-Resource Language Modelling of South African Languages.
  Stuart Mesham, Luc Hayward, Jared Shapiro and Jan Buys.
  SACAIR 2021. [Code]
  
  
• RepGraph: Visualising and Analysing Meaning Representation Graphs.
  Jaron Cohen, Roy Cohen, Edan Toledo and Jan Buys.
  EMNLP 2021 System Demonstrations. [Demo] [Code]