When the persuasive voice of Dorothy Gordon, the Director General of Ghana’s first Advanced Information Technology Institute, encourages you to organise an event in Ghana, you listen. When United Nations Secretary General Kofi Atta Annan suggests that you might help create new ties with Ghana, you feel the need to make it happen. This was the position of Mitja Jermol during a UNESCO event in 2009. His response was to develop new exciting outreach activities from PASCAL2 to Ghana.
Ghana. A country of many peoples and languages, rich in culture and natural resources. Like many African countries, it has a turbulent history, filled with bitter struggles against colonial rule, resulting in it being the first to attain independence in 1957. Its new leaders promoted pan-Africanism – unity of mind and spirit of African people worldwide. Ghana continues these ideas, teaching the principles of freedom and justice, equity and free education for all, irrespective of ethnic background, religion or creed. Its recent politics have been complex, but a stable democracy is now in place. Ghana is rich in possibilities, eager to join the world in research. In 2010 and 2011, PASCAL2 came to visit.
Following the UNESCO meeting in 2009, Mitja Jermol and Colin de la Higuera visited Ghana. They went to see the new Ghana-India Kofi Annan Centre of Excellence in ICT (AITI-KACE), in Accra. “Our main line was to believe that we were able to reproduce the highly successful Pascal boot camps in Ghana,” says de la Higuera.
From the beginning of the first PASCAL Network of Excellence, workshops, summer schools and boot camps were used to great effect. These events, organised by researchers in the PASCAL community enabled researchers to communicate their findings and to teach new students the latest cutting edge research in machine learning. The boot camps were part of a concerted Curriculum Development Programme to educate and develop young researchers in the field of machine learning. To achieve this, they focussed specifically on the goal of helping Masters level students develop skills that might enable them to pursue PhDs in machine learning. Lasting up to ten days, the boot camps were a cross between a workshop and a summer school, with introductory lectures introducing areas such as mathematics, probability, clustering and basic machine learning, more detailed lectures describing latest research, and lab sessions to encourage students to develop practical skills.
During their visit to Accra, Jermol and de la Higuera met Dorothy Gordon, the Director General of the Ghana-India Kofi Annan Centre of Excellence in ICT (AITI-KACE). Gordon is a Ghanaian who obtained degrees from the University of Ghana and the Institute of Development Studies (IDS) of the University of Sussex, UK. After graduation she joined the United Nations Development Programme. Today, her career as a specialist in international development spans over 25 years. She has been actively involved in shaping Ghana’s Information and Communications Technology scene for close to 20 years.
Gordon’s work focuses on using the opportunities that exist globally to bring development to Africa. The present direction of her work relates to how best ICT can be owned by developing countries. Her vision is to move Africa from being purely a technology consumer to becoming a technology innovator, providing overall socio-economic development. This was the reason behind the creation of the AITI-KACE – the opportunities provided by artificial intelligence were compelling. “It’s important that Ghana becomes more expert in this field,” says Gordon. “It has so much potential in terms of creating new business.”
It seemed like an ideal partnership, and the visit was a great success. Five Ghanaian students were chosen to attend the next PASCAL2 boot camp, held in Marseille, France. The boot camp ran for eight days, and covered a range of topics from kernel methods to natural language processing.
It’s important that Ghana becomes more expert in this field. It has so much potential in terms of creating new business.
Following this successful event, which had helped Ghana in a small but important way, it was decided to hold another boot camp the following year. This would be the largest boot camp ever run by PASCAL. It would be held in Accra, Ghana. Organised with the assistance of the Framework 7 EU project ACTIVE (Advanced Technologies for Knowledge-Intensive Enterprises), and the newly created Knowledge 4 All Foundation (see “PASCAL2 Legacy” article), the Ghanaian boot camp would be the first attempt of this kind at focussed outreach to Africa. It was seen as a capacity transfer boot camp, designed to help launch the Advanced Information Technology Institute in Accra. The boot camp was to be an international event, consisting of 10 days of intensive training (lectures, practical and lab sessions) and designed for master level students, researchers and lecturers interested in upgrading their knowledge and skills in the areas of Machine Learning. Gordon was enthusiastic. “This knowledge sharing forum forms the basis for further research and development collaboration between participants and the visiting AI experts,” she said at the time.
It was a unique event, bringing experts from around the world to help develop participants’ scientific skills to enable them to collaborate on research in a number of key areas: highly innovative application systems, concepts, methods and tools for Social Software and Web 2.0, Semantic Technologies, Context Mining, Context Modeling, and Context Sensitive Task
Management, Knowledge Process Mining, Knowledge Process Modeling and Pro-Active Knowledge Process Support. The event caused quite a stir in the local media, with television and newspapers reporting with some excitement that the event brought together students from universities in and outside the country, “to deliberate on the application of the artificial intelligence.”
All participants, both students and tutors, learned much from the experience. “The Ghanaian people were great and enthusiastic,” de la Higuera says. “We can put the results in two columns: pros and cons.
“Pro: we had a great boot camp, with high quality and interesting material. “Pro: we learned a lot. We learned how to outreach differently; we learned that colleagues who have so much to do will understand the importance of such actions and will be willing to help, to give their time.
“Con: our goal and their goal did not match. I understood (too late) that what mattered most for Dorothy Gordon was certification. In Ghana you don’t change jobs: you just get better conditions (or salary) in the job you have if you get brownie points. So really it would have been better to arrive saying: who wants to attend this PASCAL course, at the end of which you will get PASCAL and K4A certification? Instead, we were saying, as academics: who wants to discover this science and learn? Well, something else I had to learn.”
There can be no doubt that this unique outreach event was a great success for the participants. Longer-term collaborations inevitably take time to mature, and may become more resilient as new Ghanaian researchers in Machine Learning emerge.
“I believe strongly,” says de la Higuera, “that both from the real successes we had and the errors we made, we can, via Knowledge 4 All, propose new initiatives.”