Research @CSIR

Currently I am at the CSIR, focused on Data Science. I lead a Data Science team. In general, I am interested in topics within a wide range of application areas. With our students looking at Anomaly Detection, Social Media mining and other topics connected to Data Science and Machine Learning. Projects will be updated in time.

Team

Projects

Active projects within our group is working on:

• Project Quirk – Anomaly detection with humans in the loop.
• Project Siphon – Automated Safety And Security incident detection on Social Media.
• Project Horde – Crowdsourcing for Social Media post tagging.
• Project Motheo – A smart document analysis platform for monitoring and evaluation purposes.
• Project Khulo – Career Trajectory Prediction and Evaluation Platform.
• Project Hlangana – Graph Mining
• Project Fithla – Machine Learning on Encrypted Data

Previous Projects

• Elections Media Monitoring – Social Media Media Monitoring for the 2016 South African Municipal Elections

Publications

• Bringing Sequential Feature Explanations to life. Tshepiso Mokoena, Ofentswe Lebogo, Asive Dlaba, Vukosi Marivate [2017]
• Employment relations: A data driven analysis of job markets using online job boards and online professional networks. Web Intelligence 17,  Vukosi Marivate, Nyalleng Moorosi [2017]
• Semi-supervised probabilistics approach for normalising informal short text messages. 2017 Conference on Information Communication Technology and Society (ICTAS). Abiodun Modupe, Turgay Celik, Vukosi Marivate, Melvin Diale [2017]
• A critical and ecosystem consideration of data for sustainable development in Africa. International Conference on Social Implications of Computers in Developing Countries. Nyalleng Moorosi, Mamello Thinyane, Vukosi Marivate [2017]
• Tackling Imbalanced Data Sets For Sequential Feature Explanation Generation Using Cost Sensitive Learning And Sampling. AusDM. Tshepiso Mokoena, Vukosi Marivate, Turgay Celik [2016]
• A Multifaceted Approach to Bitcoin Fraud Detection: Global and Local Outliers. ICMLA. Patrick Monamo, Vukosi Marivate and Bhekisipho Twala [2016]
• Research using Social Media Data, Developing Country Perspectives. AfriCHI. Vukosi Marivate and Nyalleng Moorosi [2016] Workshop.
• Catching Crime: Detection of Public Safety Incidents using Social Media. PRASA. Vukosi Marivate and Pelonomi Moiloa [2016]
• Unsupervised Learning for Robust Bitcoin Fraud Detection. ISSA. Patrick Monamo, Vukosi Marivate and Bhekisipho Twala [2016]
• Extracting South African safety and security incident patterns from social media. PRASA.  Vukosi Marivate [2015]
• Privacy In Mining Crime Data From Social Media: A South African Perspective. INFOSEC. Nyalleng Moorosi and Vukosi Marivate [2015]

The Group

Data Scientists and Engineers

• Melvin Diale
• Asive Dlaba
• Vukosi Marivate
• Nyalleng Moorosi
• Tyrone Naidoo
• Mahlatse Ratsoma

Masters Students

• Yongama Feni (Wits)
• Mpho Mokoatle (NWU)
• Tshepiso Mokoena (Wits)
• Patrick Monamo (UJ)
• Koena Monyai (Wits)
• Jessica Nemasisi (Wits)

PhD Students

• Nenekazi Mkuzangwe (UJ)
• Abiodun Modupe (Wits)

Interns/Vac-Students (Always In Transit)

• See here: Interns and Vac-Students

Alumni

• Ofentswe Lebogo

Group Achievements

• 3rd Place Responsive Cities Challenge – City of Tshwane
• Egg Drop Challenge Runner Ups (whoop whoop)
• Human Capital Development award from CSIR MDS Excellence Awards.
• Organising Committee: Deep Learning Indaba
• Microsoft Azure Research Grant
• SKA Big Data Summer School
• AIMS Data Science Workshop

Further Learning

• Introduction to Apache Spark – Aug 2016
• Intro to Hadoop and MapReduce – Nov 2015 (Catching up with current tools and refresher)