v2.0.0 of my Data Cleaning Toolkit is live.
You gave feedback. I listened.
What changed:
• Input validation; no more silent crashes
• Clear error messages
• Null reporting before & after cleaning
• Coercion warnings
• Updated docs
Free. Open source. One import.
Next in our keynote announcements for #DLI2026, we're excited to welcome Prof. Vukosi Marivate @vukosi to the stage in Lagos, Nigeria! Prof. Marivate is Director of the African Institute of Data Science and Artificial Intelligence (AfriDSAI), the ABSA UP Chair of Data Science and Professor of Computer Science at @UPTuks, where he leads the @DSFSI_Research group. His research sits at the intersection of machine learning, artificial intelligence and natural language processing, with a particular focus on African and other low-resource languages. He co-founded @LelapaAI, the @MasakhaneNLP and the @DeepIndaba.
His keynote is titled "What Do Our Benchmarks Actually Measure? Evaluation Challenges for African Language AI"
Over the past decade, research on African languages in natural language processing and speech technologies has accelerated. New datasets, multilingual models and collaborative research initiatives have delivered measurable gains across machine translation, automatic speech recognition and language modelling, showing that progress for historically under-resourced languages is both possible and increasingly visible. Yet as the field begins to envision supporting thousands of the world's languages, a critical question emerges. How should progress actually be measured? Many current benchmarks were designed for well-resourced languages and assume standardised text, stable orthographies and abundant labelled data. Applied to African and other Indigenous languages, they often capture only a narrow slice of linguistic reality while privileging what is easiest to measure. This keynote explores the growing gap between advances in language modelling and the methods used to evaluate them, and argues that rethinking evaluation is essential for truly multilingual AI. Moving forward, evaluation frameworks must better reflect linguistic diversity, community priorities and the complex sociotechnical contexts in which these languages are used.
Prof. Marivate holds a PhD in Computer Science from Rutgers University, and an MSc and BSc in Electrical Engineering from the University of the Witwatersrand. He was recently honoured with the Order of Mapungubwe (Silver), one of South Africa's highest national awards, for his contributions to artificial intelligence and computer science. He also serves on the @UN Independent Scientific Panel on AI and the African AI Council at Smart Africa.
Welcome on board, Prof. Marivate!
Virtual Indaba: Registrations still open: https://t.co/d74RjD6wC6
#DLI2026 #Indaba2026
“You have thousands of moments ahead of you. The important thing isn’t to get them all right; it’s to find a way to keep moving forward.”
---- Sundar Pichai
Happy New week, just move forward.
Before dropping rows with large amount of missing values, investigate what the nan might mean...
I have to investigate 30 rows with ridiculously high amount of missing values. Happy Sunday BTW mtcheew
At Tech Sphere Academy, community is everything to us.🫶
So now it’s time to make the most of the community.
Say hello, connect with fellow members, join the conversations, and don’t be afraid to put yourself out there. Some of the best opportunities start with a simple introduction. 💙