This project explored the use of social media in Cameroon during the Anglophone Crisis and in the run-up to the elections in October 2018. I worked in partnership with r0g_, a non-profit agency for open culture and critical transformation, the United Nations and UNESCO.

Hate Speech Interface Design Overview

Background

The Anglophone Crisis refers to a period of unrest and conflict in the English-speaking regions of Cameroon, which began in 2016 and escalated leading up to the October 2018 elections. The crisis was rooted in long-standing grievances by English-speaking Cameroonians, who constitute about a fifth of the country's population, over perceived discrimination and marginalization by the Francophone-dominated government and state institutions.

In late 2016, protests by lawyers and teachers demanding greater autonomy for the Anglophone regions quickly escalated into a wider movement, with many calling for a separate state to be established. The government responded with a heavy-handed crackdown, using the military to quell protests and detain activists. This only fueled further anger and unrest, and the situation soon descended into an armed conflict, with separatist groups forming and clashing with government forces.

The situation worsened leading up to the October 2018 elections, with reports of widespread human rights abuses and violence on both sides. The elections took place amidst a climate of fear and intimidation, with many Anglophone regions effectively shut down due to the conflict. The crisis had a significant impact on the outcome of the elections, with voter turnout in the affected regions being low and the ruling party, the Cameroon People's Democratic Movement, winning a landslide victory.

The Anglophone Crisis continues to this day, with no resolution in sight. The situation has caused immense suffering for the local population, with thousands of people displaced and reports of widespread human rights abuses. The crisis has also had regional and international repercussions, with calls for a resolution to the conflict growing louder.

Social Media

Social media played a significant role in the Anglophone Crisis in Cameroon, both as a tool for activists to organize and mobilize, and as a means of spreading information and amplifying their message.

On the one hand, social media platforms such as Facebook, Twitter, and WhatsApp were used by Anglophone activists to coordinate protests, share news and updates, and build support for their cause. This allowed them to bypass traditional media outlets, which were largely controlled by the government, and get their message out to a wider audience.

On the other hand, social media also became a tool for the government to monitor and suppress dissent. Authorities used social media to track the activities of activists and separatists, and to spread propaganda and misinformation. In some cases, the government also used social media to spread false information and sow discord, in an attempt to undermine the credibility of the opposition.

Hate Speech

As the conflict escalated and tensions rose, social media platforms became a breeding ground for hateful and inflammatory language, with many users resorting to insults, personal attacks, and threats of violence.

Hate speech was used to further polarize the conflict along linguistic and ethnic lines, with individuals and groups using social media to spread hate speech against those from the other side of the conflict. This fueled resentment and mistrust, and made it even more difficult to find a peaceful resolution to the crisis.

In response to the rise of hate speech, some social media platforms, such as Facebook, took steps to remove inflammatory content and block accounts that were found to be spreading hate speech. However, these efforts have been limited, and much of the hate speech remains accessible on the platforms.

It is worth noting that the use of hate speech on social media during the Anglophone Crisis is not unique to Cameroon, and similar trends have been observed in other conflicts and crisis situations around the world. The rise of hate speech on social media highlights the need for more effective mechanisms to monitor and regulate the content on these platforms, in order to prevent the spread of hate and promote peace and understanding.

Task

My task was to collect data from particular Facebook users, news corporation pages and public persons and create an interactive data visualisation which explores what was happening in depth. The data was collected through the Facebook Graph API and included status post updates, comments, likes and shares made to that particular page/person.

Interface

The interface presents an overview of the data collected during 31 days. The X-axis shows the names of Facebook users and, beside them, their data graph displaying a combination of likes, comments and shares for that particular day. Moving forward and clicking on a bar (seen in the second image), the user is directed to a page which displays all the posts made by that Facebook page. Clicking on a box displays the contents of that post and a table which shows specific hate speech terminology used.

Data overview showing posts made by a particular user in one day.

Hate Speech Interface Profile Overview

Clicking into a post, you can then view the contents of the post, including all comments made.

Hate Speech Interface Profile Day Overview

You can filter by keywords which have been identified as hate speech.

Hate Speech Interface Profile Day Filtered
Hate Speech Interface Profile Day Filtered

This application is built with HTML, CSS and JavaScript. I experimented with d3.js and produced some outcomes for this project; however, using JavaScript was the way forward for what I wanted to achieve.

Hate Speech Data Collected Overview

The image above shows how I organised the data for each Facebook profile after the data was collected with FacePager. The Excel files contain all the posts made within the timeframe from 15 June to 15 August 2018. All the comments made by users and the number of likes and shares are shown too.