KnowledgE ENgineering team
Funded projects - Recent highlights - Members and collaboratorsThe aim of the team's activities is to contribute computing theory, methods, and techniques to the knowledge society.
The scope of the KEEN team is knowledge engineering in its broad sense. This includes ontology engineering, the Semantic Web, intelligent (logic-based, ontology-driven) conceptual modelling, and natural language generation.
Research projects
- MoRe NL: foundations of a Modular Realisation Engine for Nguni Languages, funded by the NRF (2020-2022)
- ALMA – Modeling of Knowledge Networks in Medieval Romance Speaking Europe Based on Linguistic Data, led by the Heidelberg Academy of Sciences and Humanities, funded by the German Akademienprogramm. Aug 2022 - 2025.
- QueST: Querying Smart Text, approved for funding by SADiLaR.
- Various funding sources over the past years; mainly: NRF, HPI Research School in Computer Science for Africa (CS4A) at UCT, DOT4D@UCT, ULPDO@UKZN.
- Recently completed projects:
- Teaching material development grant for the "An introduction to Ontology Engineering" textbook, funded by the Digital Open Textbooks for Development project at UCT. Project duration: 2019-2020 (1 year + top-up funding in 2021).
- Erasmus+ funding on ontologies and machine learning with Poznan University of Technology (2020-2022), where I was supposed to be PI with Agnieszka Lawrynowicz and do exchanges on that topic, but the colonial admin of PUT preferred otherwise.
- ARISTOTELES: Methodology and algorithms for automatic revision of ontologies in task based scenarios. Funded by the Polish National Science Center (2015-2018) with PI Agnieszka Lawrynowicz.
- Medical Machine Translation, in collaboration with the University of Edinburgh, UK. Funded by the EPSRC, UK (2017-2018).
- GeNi: A Grammar Engine for Nguni natural language Interfaces. Funded by the NRF (2015-2017).
- Foundations for Nguni natural language generation systems. UCT Science Faculty Launching grant (2014-2017).
Recent highlights
- A journal paper about a surface realisation architecture for low-resourced African languages, with as main author Zola' Mahlaza and Maria Keet, which is in print with the ACM Transactions on Asian and Low-Resource Language Information Processing journal.
- A conference paper about an architecture for generating educational questions from ontologies, accepted for the MTSR'22 conference, with Toky Raboanary and Maria Keet.
- A keynote talk at the Bio-Ontologies COSI at ISMB'22 on Encoding biases' influences on development and use of ontologies in the life sciences, with slides.
- A Digital Humanities Colloquium at SADiLaR by Maria Keet, on Natural Language Generation for Agglutinating African Languages, with a recording and slides
- A paper about multilingualism in bio-ontologies, accepted for SWAT4HCLS'22, with Frances Gillis-Webber and Maria Keet.
- Conference organisation of the ISWC'22 resources track, and recently also IJCAI 2021 Demonstrations track, the 22nd International Conference on Knowledge Engineering and Knowledge Management (EKAW 2020), ISWC 2019 Outrageous Ideas Track, and others.
Members and collaborators
Leader
|
Maria Keet |
Postgraduate students
|
Mary-Jane Antia (PhD; 2019-) area: ontology engineering, competency questions |
|
Frances Gillis-Webber (PhD; 2019-) area: multilingual ontologies |
|
Toky Raboanary (PhD; 2019-) area: ontology engineering |
|
Zola Mahlaza (PhD; 2018-2022, graduation in Dec'22) topic: isiXhosa and isiZulu NLG |
|
Leighton Dawson (MSc, 2020-) topic: ODPs |
|
Crystal Yang (M. IT; 2022-) topic: knowledge management in legal IT |
|
Jacques de Lange (M. IT; 2021-) topic: NLP with SA English |
|
Kyle Seakgwa (M. IT; 2021-) topic: knowledge extraction for artificial moral agents |
BSc honours students
- TBA for 2023
- Sindiso Mkhatswa
- Toky Raboanary
- Universidad Nacional del Sur, Bahia Blanca, Argentina - Depto. Ciencias e Ingeniería de la Computación (Pablo Fillottrani)
- Poznan University of Technology, Poland - Laboratory of Operational Research and Artificial Intelligence (Agnieszka Lawrynowicz)
- South African Centre for Digital Language Resources SADiLaR (Langa Khumalo)
- University of Waterloo, Canada - Computer Science department (David Toman)
- Council for Scientific and Industrial Research (CSIR), South Africa (Zubeida Khan)
- Digital Libraries Lab, UCT (Hussein Suleman)
- School of Informatics, University of Edinburgh, UK, (Kenneth Heafield, Alexandra Birch)
- Joan Byamugisha (PhD; 2015-2019); topic: Runyankore knowledge-to-text NLG for healthcare
- Yamiko Msosa (PhD; 2014-2018); topic: semantics-driven linincal practice guideline changes
- Zubeida Khan (PhD; 2014-2017); topic: ontology modularity
- Zola Mahlaza (MSc; 2016-2017); topic: data-to-text NLG for weather bulletins in isiXhosa and isiZulu
- Cameron Kyle (M. IT; 2017); topic: CNLs for finance
- Ana Dauane (2021; BSc honours): SALSAEditor: Creation of a Domain Specific Modelling Language for Latin Dances
- James Kriel (2021; BSc honours): Video Annotation as a Tool to Supplement in-person DanceClasses
- Junior Moraba (2021; BSc honours): Development of a Finance Based IsiZulu NLG System that Verbalises Numbers in Context
- Amy Solomons (2021; BSc honours): A Personal Finance Management Mobile Application Incorporating NLG
- Kouthar Dollie (2020; BSc honours): Social Dance App -- MetaDance
- Carryn Joseph (2020; BSc honours): Social Dance App -- Dancebase
- Alec Badenhorst (2020; BSc honours): CASEText: A computational analysis of SA English
- Umar Khan (2020; BSc honours): CASEText: corpora for SA English
- Mandisa Baleni (2020; BSc honours): KnowDat: Materialisation of deductions (as part of KnowID)
- Bradley Malgas (2020; BSc honours): KnowDat: ggraphical query interface (as part of KnowID)
- Alka Baijnath (2019; BSc honous): Salsational Dance Application
- Jordy Chetty (2019; BSc honous): Salsational Dance Application
- Micara Marajh (2019; BSc honous): Salsational Dance Application
- Kyle Robbertze (2019; BSc honous): ontology-annotated smart textbooks
- Steve Wang (2019; BSc honous): ontology-based question generation
- Jarryd Dunn (2019; BSc honous): comparing end-to-end models and templates for Generating text (main supervisor: Zola Mahlaza)
- Matthew Poulter (2019; BSc honous): comparing end-to-end models and templates for Generating text (main supervisor: Zola Mahlaza)
- Nthabiseng Mashiane (2017; BSc honous); topic: spelling error detection for isiXhosa
- Frida Mjaria (2017; BSc honous); topic: spelling error correction for isiZulu
- Siseko Neti (2017; BSc honous); topic: spelling error detection for isiXhosa
- Nikhil Gilbert (2017; BSc honous); topic: Automated marker for language learning exercises
- Soham Singh (2017; BSc honous); topic: difficulty levels with the automated marker for langauge learning
- Kgotso Nkosi (2017; BSc honous)
- Ameerah Allie (2016; BSc honous); topic: Test-driven development of ontologies
- Kieren Davies (2016; BSc honours); topic: Toward Test-driven development of ontologies
- Chiadika Emeruem (2021, RA/programmer): OE tools and tutorial
- Steve Wang (2021, RA/programmer): OE tools and tutorial
- Junior Moraba (2021, RA/programmer): MoReNL project, on isiZulu numbers
- Sindiso Mkhatshwa (2021, programmer): MoReNL project, ToCT tool
- Kouthar Dollie (2020; research assistant): MorReNL project
- Blessed Chitamba (2020; scientific programmer): MorReNL project
- Gerald Ngumbulu (2020; scientific programmer): MorReNL project
- Zola Mahlaza (2019; RA): OE tutorial
- Frances Gillis-Webber (2019; RA): OE tutorial
- Michael Harrison (2019; scientific programmer): OE tools
- Toky Raboanary (2019; scientific programmer): OE tools
- Stephan Jamieson (2019; scientific programmer): KnowID
- Zola Mahlaza (2017; scientific programmer); on the isiZulu verb grammar and medical machine translation
- Siseko Neti (2017; scientific programmer); isiZulu grammar and spellchecker
- Norman Pilusa (2016, 2017; scientific programmer); isiZulu spellchecker
- Takunda Chirema (2016; scientific programmer); a tool for modelling with relations and roles
- Musa Xakaza (2016; scientific programmer); linking the verbaliser to Owlready