From monitoring the country’s coastline for illegal fishing to assessing the needs of the public education system, these and other problems facing the country could be addressed by using data-science solutions. Several students, from universities across the country showcased their data-science projects funded by the Department of Science and Technology (DST) this week.
The Data Science for Impact and Decision Enhancement (DSIDE) programme funded by the DST and implemented by the Council for Scientific and Industrial Research (CSIR) is aimed at building capacity in the field of data science and to bring solutions to real-world problems.
The programme is jointly hosted at the CSIR by two (2) business units, namely the Meraka Institute and the Modelling and Digital Science (MDS). The aim of the programme is to support capacity building in the ever growing field of data science by scheduling recruits to participate in mentor-guided and learn-by-doing problem solving of real world needs as presented by different stakeholders.
The DSIDE programme puts emphasis on problem solving, creativity, and encourages a curious mind from the students. Experienced mentors from the CSIR data science community introduce machine learning topics, tools, theories, and guide students in this project-driven environment. The students were selected after an intensive recruitment exercise from tertiary institutions all over the country.
The DSIDE Programme Manager, Dhiren Seetharam, says the focus of this initiative is to ensure that students learn by solving real problems the country is grappling with like analyzing data on crime, looking at the ongoing drought, illegal fishing, threats to security, and building maintenance, among others.
Macdonald Maringa, an Actuarial Science student at Pretoria University, Omolemo Matlou from University of Johannesburg, and Langa Khahla, a graduate in Statistical Mathematics, have developed an application called Fundza Platform. The application was used to analyze the 2014 Grade 12 results. The application allowed its users to assess how the lack of infrastructure at schools impacted the pass rate countrywide.
Fundza Platform was developed as a tool to assist the Department of Basic Education to make informed decisions on how to manage its resources across the country. Public education forms a large part of the national budget, but learners continue to perform poorly in the national matric exams.
Analyzing the data obtained from the 2014 matric results, Maringa’s group found no distinct relationship between infrastructure and the pass rate, but schools with adequate infrastructure usually performed better.
For example, Godide High in Uthungulu in KwaZulu-Natal, obtained a 29% matric pass rate in 2014, with classrooms as its only infrastructure without facilities like a library, while Protea Glen High School in Soweto obtained a 93% pass rate, with classrooms and administration building as its only facilities.
Also in Uthungulu, Zakhekahle high obtained a 65,6% pass rate and was receiving nutrition through the government’s school nutrition programme – the only difference between it and Godide High.
“We have a lot of factors that can demonstrate a relationship between infrastructure and the pass rate really exists; such as learner and teacher ratio, quintiles and geographic location,” said Maringa.
As the students analysed data obtained from 2014 only, Maringa believes that the project needs to look further and get more data to perform time-series analysis, looking at school performance over at least 10 years.
Another significant application is "Project Mati", which uses data on the drought to assist the Department of Water and Sanitation (DWS) to develop an information system on water-related matters in the country. Project Mati also aims to develop a web application that will help the Department to best display the data, mapping the affected areas, settlements and municipalities.
This application assists the DWS to deploy its resources to areas most vulnerable to the drought.
Monitoring the country’s coastline, Project-Sea Far is a project tasked for monitoring the South African oceans mainly the fishing activities. This includes SA and Non-SA vessels that traffic through South Africa’s Exclusive Economic Zone (EEZ), the main objective centered around vessels that frequent regularly, their activities and also their legitimacy.
With a lot of cases of illegal fishing the Project-Sea Far system aims to monitor such activities, visualise information and aid in reducing illegal fishing. South Africa’s EEZ consists of 1,535,538 km2 where the exploration and exploitation of marine resources is authorised exclusively for the South African continent.
This project enables the relevant authorities to understand fishing events and activities within maritime environment and can be enhanced by the automatic identification and classification of illegal fishing vessels activities.
Project-Sea Far designed a software system that is able to monitor South Africa water’s finishing activities. This was achieved by extracting relevant Automatic Identification System AIS data to track ships movement, speed and their destination in South African waters.