Siyabulela “Siya” Sokomani, a young entrepreneur from South Africa, runs a commercial nursery that grows thousands of native saplings every year. By nurturing those trees, he is restoring degraded land and fighting climate change in his community’s township in Cape Town. He’s one of countless African leaders critical to achieving the African Forest Landscape Restoration Initiative’s (AFR100) goal of building vital landscapes across 100 million hectares by 2030.

Yet historically, people like Sokomani have been underrepresented at major restoration events and on social media. The Global Landscapes Forum (GLF), the world’s largest platform for sharing information on sustainably managing land, is traditionally held in Bonn, Germany. This can make participation nearly impossible for people like Sokomani who face structural obstacles with visas or travel expenses. Voices like Sokomani’s also get drowned out on social media by an international development community that often disregards marginalized voices. As a result, key discussions often happen without the on-the-ground perspectives vital to effectively restore the world’s degraded landscapes.
A New Approach

Recognizing the need for more inclusivity, the GLF decided to host its first major event in Africa. It first came to Nairobi, Kenya in 2018, followed by a second event in Accra, Ghana in 2019. WRI analyzed the impact that the change of continent had on participation from African restoration leaders.

This analysis found that hosting the GLF in Africa, both in Kenya and Ghana, resulted in a more diverse audience — half of the 13,000 people who attended the 2018 event, both virtually and in person, were from African countries. The diversity of voices on social media was even more impressive. The 2018 GLF in Nairobi reached 52 million people on social media, driven in large part by African restoration leaders.

The paper detailing this analysis, which was published in the Journal of Computational Social Science, establishes a quantitative method to measure the engagement of African voices during important environmental events. We used machine learning, natural language processing and computer vision techniques to analyze millions of tweets. After anonymizing the data, we identified the locations, age group, gender and ethnicity of each user. Next, we used algorithms — similar to those that power Google searches — to read through and categorize every tweet. This allowed us to analyze how people with different locations, ages, genders and ethnicities interacted with the GLF.

Here’s what we learned ...read further from the source above or the whole study
"Identifying social media user demographics and topic diversity with computational social science: a case study of a major international policy forum"