Publications

2025

1. Identification of Social Media Users that Perpetuate Xenophobic Attitudes and Hate Speech Narratives in South Africa

   Carl Plessis, Michael Plessis, Koena Ronny Mabokela, and 2 more authors

   In Southern African Conference for Artificial Intelligence Research, 2025
2. Mafoko: Structuring and Building Open Multilingual Terminologies for South African NLP

   Vukosi Marivate, Isheanesu Dzingirai, Fiskani Banda, and 8 more authors

   arXiv preprint arXiv:2508.03529, 2025
3. Cross-lingual embedding methods and applications: A systematic review for low-resourced scenarios

   Thapelo Sindane, Vukosi Marivate, and Abiodun Modupe

   Natural Language Processing Journal, 2025

   Abs DOI

   The field of Natural Language Processing (NLP) has achieved significant success in various areas, such as developing large-scale datasets, algorithmic complexity, optimized computing capabilities, refined individual and community expertise, and more, particularly in languages such as English, French, and Spanish. However, such global north unilateral strides have inadvertently created a substantial representation bias towards many languages categorized as low-resourced languages, with the majority being African languages. As a result, rudimentary resources such as stopwords, lemmatizers, stemmers, and word embeddings, as well as advanced multilingual transformer-based models remain under-developed for these languages. Compounding these circumstances is the lack of insights surrounding the development of these resources in the low-resourced context (e.g., how to develop embeddings for morphologically rich languages). Looking back, research priorities aiming to create these resources, largely motivated by the high cost attached to remedying these issues shifted, leading to the rise of alternative methods such as cross-lingual transfer learning (CLTL). CLTL involves transferring domain knowledge gained from supervised training to a domain with limited supervision signals. This study conducts a systematic literature review of CLTL techniques, in the context of cross-lingual models and embeddings, looking at their mathematical foundations, application domains, evaluation metrics, languages covered, and the latest developments. The findings of this study offer valuable insights into the present scenario of CLTL techniques, identifying areas for future research and development to advance cross-lingual natural language processing applications specifically in low-resourced settings.

2024

1. ChatGPT as a Text Annotation Tool to Evaluate Sentiment Analysis on South African Financial Institutions

   Miehleketo Mathebula, Abiodun Modupe, and Vukosi Marivate

   IEEE Access, 2024

   DOI
2. Fine-Tuning Retrieval-Augmented Generation with an Auto-Regressive Language Model for Sentiment Analysis in Financial Reviews

   Miehleketo Mathebula, Abiodun Modupe, and Vukosi Marivate

   Applied Sciences, 2024

   Abs DOI

   Sentiment analysis is a well-known task that has been used to analyse customer feedback reviews and media headlines to detect the sentimental personality or polarisation of a given text. With the growth of social media and other online platforms, like Twitter (now branded as X), Facebook, blogs, and others, it has been used in the investment community to monitor customer feedback, reviews, and news headlines about financial institutions’ products and services to ensure business success and prioritise aspects of customer relationship management. Supervised learning algorithms have been popularly employed for this task, but the performance of these models has been compromised due to the brevity of the content and the presence of idiomatic expressions, sound imitations, and abbreviations. Additionally, the pre-training of a larger language model (PTLM) struggles to capture bidirectional contextual knowledge learnt through word dependency because the sentence-level representation fails to take broad features into account. We develop a novel structure called language feature extraction and adaptation for reviews (LFEAR), an advanced natural language model that amalgamates retrieval-augmented generation (RAG) with a conversation format for an auto-regressive fine-tuning model (ARFT). This helps to overcome the limitations of lexicon-based tools and the reliance on pre-defined sentiment lexicons, which may not fully capture the range of sentiments in natural language and address questions on various topics and tasks. LFEAR is fine-tuned on Hellopeter reviews that incorporate industry-specific contextual information retrieval to show resilience and flexibility for various tasks, including analysing sentiments in reviews of restaurants, movies, politics, and financial products. The proposed model achieved an average precision score of 98.45%, answer correctness of 93.85%, and context precision of 97.69% based on Retrieval-Augmented Generation Assessment (RAGAS) metrics. The LFEAR model is effective in conducting sentiment analysis across various domains due to its adaptability and scalable inference mechanism. It considers unique language characteristics and patterns in specific domains to ensure accurate sentiment annotation. This is particularly beneficial for individuals in the financial sector, such as investors and institutions, including those listed on the Johannesburg Stock Exchange (JSE), which is the primary stock exchange in South Africa and plays a significant role in the country’s financial market. Future initiatives will focus on incorporating a wider range of data sources and improving the system’s ability to express nuanced sentiments effectively, enhancing its usefulness in diverse real-world scenarios.

3. Misinformation Detection: A Review for High and Low-Resource Languages

   Seani Rananga, Bassey Isong, Abiodun Modupe, and 1 more author

   Journal of Information Systems and Informatics, Dec 2024

   DOI

2023

1. Multimodal Misinformation Detection in a South African Social Media Environment

   Amica De Jager, Vukosi Marivate, and Abiodun Modupe

   In Southern African Conference for Artificial Intelligence Research, Dec 2023
2. Integrating Bidirectional Long Short-Term Memory with Subword Embedding for Authorship Attribution

   Abiodun Modupe, Turgay Celik, Vukosi Marivate, and 1 more author

   In 2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC), Dec 2023

   DOI
3. Financial Inclusion in Sub-Saharan Emerging Markets: The Application of Deep Learning to Improve Determinants

   Johnson S Dlamini, Linda Marshall, and Abiodun Modupe

   In Southern African Conference for Artificial Intelligence Research, Dec 2023

2022

1. Post-Authorship Attribution Using Regularized Deep Neural Network

   Abiodun Modupe, Turgay Celik, Vukosi Marivate, and 1 more author

   Applied Sciences, Dec 2022

   Abs DOI

   Post-authorship attribution is a scientific process of using stylometric features to identify the genuine writer of an online text snippet such as an email, blog, forum post, or chat log. It has useful applications in manifold domains, for instance, in a verification process to proactively detect misogynistic, misandrist, xenophobic, and abusive posts on the internet or social networks. The process assumes that texts can be characterized by sequences of words that agglutinate the functional and content lyrics of a writer. However, defining an appropriate characterization of text to capture the unique writing style of an author is a complex endeavor in the discipline of computational linguistics. Moreover, posts are typically short texts with obfuscating vocabularies that might impact the accuracy of authorship attribution. The vocabularies include idioms, onomatopoeias, homophones, phonemes, synonyms, acronyms, anaphora, and polysemy. The method of the regularized deep neural network (RDNN) is introduced in this paper to circumvent the intrinsic challenges of post-authorship attribution. It is based on a convolutional neural network, bidirectional long short-term memory encoder, and distributed highway network. The neural network was used to extract lexical stylometric features that are fed into the bidirectional encoder to extract a syntactic feature-vector representation. The feature vector was then supplied as input to the distributed high networks for regularization to minimize the network-generalization error. The regularized feature vector was ultimately passed to the bidirectional decoder to learn the writing style of an author. The feature-classification layer consists of a fully connected network and a SoftMax function to make the prediction. The RDNN method was tested against thirteen state-of-the-art methods using four benchmark experimental datasets to validate its performance. Experimental results have demonstrated the effectiveness of the method when compared to the existing state-of-the-art methods on three datasets while producing comparable results on one dataset.

2021

1. Call Centre Shift Schedule Optimisation using Local Search Heuristics

   Liketso Nthimo, Tshepiso Mokoena, Abiodun Modupe, and 1 more author

   In 2021 IEEE AFRICON, Dec 2021

   DOI
2. Investigating Statistical and Machine Learning Techniques to Improve the Credit Approval Process in Developing Countries

   Moses Lusinga, Tshepiso Mokoena, Abiodun Modupe, and 1 more author

   In 2021 IEEE AFRICON, Dec 2021

   DOI

2020

1. Investigating an Approach for Low Resource Language Dataset Creation, Curation and Classification: Setswana and Sepedi

   Vukosi Marivate, Tshephisho Sefara, Vongani Chabalala, and 4 more authors

   In Proceedings of the first workshop on Resources for African Indigenous Languages, Dec 2020
2. Low resource language dataset creation, curation and classification: Setswana and Sepedi

   Vukosi Marivate, Tshephisho Sefara, Vongani Chabalala, and 4 more authors

   arXiv preprint arXiv:2004.13842, Dec 2020
3. Low resource language dataset creation, curation and classification: Setswana and Sepedi

   Vukosi Marivate, Tshephisho Sefara, Vongani Chabalala, and 4 more authors

   arXiv preprint arXiv:2004.13842, Dec 2020

2019

1. Yorùbá Gender Recognition from Speech Using Neural Networks

   Tshephisho Joseph Sefara and Abiodun Modupe

   In 2019 6th International Conference on Soft Computing & Machine Intelligence (ISCMI), Dec 2019

   DOI

2016

1. Feature selection and support vector machine hyper-parameter optimisation for spam detection

   Melvin Diale, Christiaan Van Der Walt, Turgay Celik, and 1 more author

   In 2016 Pattern Recognition Association of South Africa and Robotics and Mechatronics International Conference (PRASA-RobMech), Dec 2016

   DOI

2015

1. Development and evaluation of a system for normalizing Internet slangs in social media texts

   Adedoja A Adedamola, Abiodun Modupe, and Olumuyiwa J Dehinbo

   In Proceedings of the World Congress on Engineering and Computer Science, Dec 2015

2014

1. Filtering of mobile short messaging service communication using latent Dirichlet allocation with social network analysis

   Abiodun Modupe, Oludayo O Olugbara, and Sunday O Ojo

   In Transactions on Engineering Technologies: Special Volume of the World Congress on Engineering 2013, Dec 2014

2013

1. Experimental Comparison of Genetic Algorithm and Ant Colony Optimization to Minimize Energy in Ad-hoc Wireless Networks

   Ibukunola A Modupe, Oludayo O Olugbara, Sunday O Ojo, and 1 more author

   In Proceedings of the World Congress on Engineering and Computer Science, Dec 2013
2. Minimizing Energy Consumption in Wireless Ad hoc Networks with Meta heuristics

   Ibukunola. A. Modupe, Oludayo. O. Olugbara, and Abiodun. Modupe

   Procedia Computer Science, Dec 2013

   The 4th International Conference on Ambient Systems, Networks and Technologies (ANT 2013), the 3rd International Conference on Sustainable Energy Information Technology (SEIT-2013)

   Abs DOI

   The objective of this study is to describe an energy function model base on Geographic Adaptive Fidelity (GAF), which is one of the best known topology management schemes used in saving energy consumption in ad-hoc wireless networks. In wireless ad-hoc network, the nodes responsible for the transmission of data are battery-operated and as a result, there is a need for energy to be conserved in order to prolong the battery lifespan. Genetic Algorithm (GA) and Simulated Annealing (SA) metaheuristics are compared to minimize the energy consumption in ad-hoc wireless networks modelled by rectangular GAF. Results show that GA and SA meta-heuristics are useful optimization techniques for minimizing the energy consumption in ad-hoc wireless networks.

3. Investigating topic models for mobile short messaging service communication filtering

   Abiodun Modupe, Oludayo O Olugbara, and Sunday O Ojo

   In Proceedings of the world congress on engineering, Dec 2013
4. Experimental Comparison of Genetic Algorithm and Ant Colony Optimization to Minimize Energy in Ad-hoc Wireless Networks

   Ibukunola A Modupe, Oludayo O Olugbara, Sunday O Ojo, and 1 more author

   In Proceedings of the World Congress on Engineering and Computer Science, Dec 2013

2012

1. Comparing supervised learning classifiers to detect advanced fee fraud activities on internet

   Abiodun Modupe, Oludayo O Olugbara, and Sunday O Ojo

   In International Conference on Computer Science and Information Technology, Dec 2012

2011

1. Exploring Support Vector Machines and Random Forests to Detect Advanced Fee Fraud Activities on Internet

   Abiodun Modupe, Oludayo O. Olugbara, and Sunday O. Ojo

   In 2011 IEEE 11th International Conference on Data Mining Workshops, Dec 2011

   DOI