AI for Social Good — Nigeria

Eniola Alese
ExplainingML
Published in
5 min readJul 13, 2018
Photo by Victor Nnakwe on Unsplash

Over the past few years, there has been so much buzz around “Artificial Intelligence, Deep Learning, and Machine Learning” and to an extent rightfully so, as advancements in this field has transformed various industries and will continue to do so in the future. However, most of these developments are applied in solving problems that occur in developed countries and unfortunately very few translate towards problems in developing countries.

“Leave this world a little better than you found it.” ― Robert Baden-Powell

I was born and spent most of my life in Nigeria and while growing up, I witnessed so many social issues plague and stunt our progress as a nation and this constantly made me sad and upset. As an adult, I have been privileged to gain an education and I’m currently pursuing a career using Machine Intelligence to develop solutions that help people. Although, machine learning isn't a silver bullet and cannot solve all of our problems, I believe it sure can help alleviate some of them.

In this post, I highlight some opportunities in which we can apply advances in this amazing field to tackle social issues in Nigeria like healthcare, unemployment and security.

Access to Healthcare

The healthcare system in Nigeria faces a lot of problems but one of its major issues is the alarming number of medical professionals in the country. Currently, we have a 1:4,000 doctor-to-patient ratio and an even worse ratio for nurses; this is due to the migration of these professionals to other countries in pursuit of better working conditions, which ultimately leaves the health needs of the population at a great loss. This problem is even worse in rural areas as most of the medical staff left, are disproportionally concentrated in major cities.

https://www.thisdaylive.com/index.php/2018/06/21/stakeholders-to-address-lack-of-sustainable-healthcare-in-nigeria/

While hoping the Nigerian government acts swiftly to fix the healthcare system and retain its talent, and also understanding that currently no technology can replace the physical and human care that doctors and nurses provide, I believe there is an immediate health gap that an AI driven solution could help fill. For example, a Telehealth platform could be developed; which provides free and on-demand medical advice about possible health conditions and also helps direct the user to the most appropriate level of care they might need . The platform could be deployed via a mobile application or via USSD (text messaging) and could leverage:

  • transfer learning from pre-trained medical knowledge models gotten from existing healthcare applications like Babylon , Babyl, Ada,
  • fine-tuning these models to learn about health issues specific to countries like Nigeria such as malaria, HIV/AIDS, tuberculosis, maternal & child mortality etc.
  • dialogue and natural language understanding to translate all ready existing medical advice from English language into local languages like Pidgin, Yoruba, Igbo, Hausa etc.

Youth Employment

Another critical social issue facing the country is the high youth unemployment rate, according to the Nigerian Bureau of Statistics 40.6% of youths within the ages (15–34) are unemployed, while 58.6% of youths within this age bracket are underemployed, and this number is steadily rising.

Yearly, about 200,000 students graduate from the university with a degree but a huge percent are unable to find suitable employment most times because the demand for jobs far outweigh its supply, or they lack the necessary skills required to pursue certain professions, or sometimes they are simply clueless on what careers to pursue, and its in this space that an AI driven solution could be utilized to ease this problem.

For instance, a skills-to-job matching platform could be used to help youths who are looking to leverage opportunities on the internet and get into new-collar freelance jobs like content writing, app development, graphics design, social media marketing, cryptocurrency trading, etc. It could employ traditional machine learning approaches like collaborative or content-based filtering to analyze information about a user’s interests, background, skills, goals etc and use this information to recommend careers that might be a strong fit to their profile and also guide the user through a personalized training syllabus to gain the required skills.

Defending Against Terror

Violent and terror based attacks on human life have almost become commonplace, and this issue is prevalent in the northeastern parts of the country, leading to the absurd loss of life, kidnappings, internal displacement of people and constant fear of terror in the hearts of Nigerians.

https://www.cfr.org/nigeria/nigeria-security-tracker/p29483

Currently, the Nigerian military along with civilian volunteers are working round the clock to defend against these attacks, one of the ways they do this is by patrolling affected cities and setting up road checkpoints to search for signs of terror activities.

While patrolling may be really helpful in preventing attacks, determining the most efficient patrol route & road checkpoint might prove to be difficult for the troops, due to the multiple variables they may need to consider. In this situation, a machine learning model could be useful as it could help generate predictions of the most likely areas an attack might occur, and also recommend optimal patrol routes and road checkpoints.

The above are just a slice of the societal challenges we face as a nation and some suggestions in trying to solve them. However, while technology offers tremendous opportunities to transform and make life better for people, there are still a number of concerns that may need to be addressed in order to successfully implement these solutions in Nigeria, such as: gathering data required to train models, lack of talent ( human-centred designers, data scientists, ML engineers), inadequate support from the government due to corruption etc.

In conclusion, I believe Machine Intelligence is here to stay and if built with the benefit of everyone in mind could have one of the greatest impact to humans in the coming decades.

--

--