Anatomy of a Hashtag

Microblogging social media websites like Twitter are data troves, but how can this information be leveraged to capture meaningful insights about humanitarian crises? This project seeks to use digital humanities tools and research methods for a close examination of a popular hashtag called #EndSARS, which is commonly used on Twitter to spread information about the mass protests against police brutality in Nigeria. Resources like  Documenting the Now  have laid out an ethical framework for the use of social media to record historical events, which has helped guide the collection of data that was used to construct this project from user-generated content on Twitter (Documenting the Now, n.d.). A dataset was curated by using open source RStudio packages that work with Twitter’s API to scrape Tweets along with their bountiful metadata. A total of 9,944 Tweets from the #EndSARS hashtag from January 19th through January 24th were scraped from the platform. Guobin Yang describes the underpinnings of digital movements in his Narrative Agency in Hashtag Activism: The Case of #BlackLivesMatter, which recognizes the versatility of this form of digital activism and the incredible power that comes from its narrative agency and salience for particular communities. Using popular digital humanities tools including Voyant, Tableau, TopicModelingTool, and web mapping software like ArcGIS, I processed the data from the hashtag using research methods like text analysis and GIS to formulate research questions and investigate how this information is processed and disseminated by social media users. Various data visualizations were created to explore the dataset, extract insights, and develop research questions that could enhance a user’s understanding of the #EndSARS hashtag and the movement itself. After sifting through the dataset and making key findings, I took this information and presented it using a digital map format with ArcGIS StoryMaps. I produced a StoryMap that strings together a narrative about the people, places, and notable events related to the #EndSARS movement in response to the Nigerian police force unit called the Special Anti-Robbery Squad.

Voyant is a powerful web-based text analysis tool that allows you to pull analytics from a text corpus through distant reading. I plugged the Tweets from my dataset into this tool so that I could approach its contents from a birds-eye view, as opposed to doing a close reading of every Tweet individually. After removing stop words from each Tweet, this tool can produce word clouds and other visualizations to enhance our ability to analyze them all together. These visualizations help surface words that are used frequently, which can help identify key terms, people, or events related to the #EndSARS movement. Although Voyant is incapable of providing the context for how these terms are being used in conversation, it gives us a jumping-off point for terms that may warrant further investigation. By looking at a word cloud of the top 45 most frequently used words, a researcher could map out these terms into key places, people, and events that could be helpful to investigate further. This may help answer research questions about who, what, when, where, why, and how.

• Nigeria: as the number one word, appearing 1,357 times throughout the dataset, we can quickly assess that the #EndSARS movement is centralized in this country or primarily affects people who are living in Nigeria
• Lagos: more specifically we notice that Lagos, the number one populous city in Nigeria and second largest city in Africa, plays a large role in the movement somehow

• Protestors: the second most common word we see from the visualization is protest and protestors, which helps us identify at least one group that is involved in the movement
• Government: along with “endbadgovernanceinnigeria” we quickly see that the #EndSARS hashtag is a political conflict that somehow involves the Nigerian government 
• Police: phrases including police, such as “endpolicebrutality” and “policebrutality” and indicates that Nigerian police are another major group that plays a role in the movement
• Fulani: also known as Fula or Fulɓe people, these are one of the largest ethnic groups in West Africa and are mostly concentrated in Nigeria
• Mbuhari: terms like “mbuhari” or “buhari” most likely refer to Muhammadu Buhari who has been serving as the President of Nigeria since 2015
• Davido: David Adedeji Adeleke (popularly known as Davido) is a Nigerian-American musician and very influential artist in Africa
• Jidesanwoolu: although this term isn’t very obvious at first, this is the Twitter handle for the official account of Babajide Sanwo-Olu, who is the Executive Governor of Lago State in Nigeria
• Herdsmen: Fulani herdsmen or Fulani pastoralists raise livestock in the Sahel and semi-arid parts of West Africa 
• Igboho: could refer to Igboho, a town in Nigeria, or the popular human rights activist Sunday Igboho who was nicknamed after his hometown in Oke-Ogun, Oyo State

• Lekkimassacre: On October 20th, 2020, the Nigerian Army violently opened fire on peaceful #EndSARS protestors at the Lekki tollgate in Lagos State, Nigeria  
• Inaugurationday: this probably appears as a frequently used term because of the inauguration of Joe Biden as the 46th President of the United States on January 20th, which occurred during the period that the dataset was created

Next, I will shift into using Tableau to generate some dashboards with simple data visualizations about the users who wrote the Tweets and the impact they had on other users. Every Tweet in the dataset comes with an abundance of metadata about the Tweet itself and the author who wrote it. This data can characterize Tweets by showing: how many likes and retweets it received, who authored it, where it was Tweeted from, and how many followers the author has. By creating a series of pie charts, bar graphs, treemaps, and bubble maps this data can be visualized to show the varying level of involvement of different users who are participating in the hashtag and how impressionable the content they generate is.

As the Voyant and Tableau visualizations pointed out, Nigeria is the focal point of the #EndSARS movement and Nigerians are currently churning out the most content for the hashtag campaign. Specifically, the hashtag has gained a lot of momentum in the city of Lagos, where an overwhelming majority of users are Tweeting from. The Voyant word clouds also identified three important groups of people from the dataset, which includes protestors, police, and government. Without context of how these terms are used in conversation, it can be difficult to situate how these groups of people are related and interacting with one another. Fortunately, the Tableau data visualizations were able to pull out which news outlets and citizen journalists are actively reporting on the movement. The treemap visualizations indicated that Al Jazeera English was one of the larger news reporters who were participating in the #EndSARS hashtag, so it could be helpful to refer back to the table of Tweets for more information in context.

Now that the movement’s origin and the power dynamic between all of the different groups involved in the #EndSARS campaign is evident, it is worthwhile to investigate the significance of the Lekki Massacre that Voyant identified from the Tweets. I created a map that shows how users are engaging in conversation about the incident by using various geolocated Tweets that reference it by name. The map consists of 34 Tweets written by various users who mention something about the events that transpired. Many users express their grief by commemorating its 3 month anniversary, which occurred over the span of the Tweets that were curated between January 19th through the 24th. According to several users like  @ReubenOdigie  and  @oga_comedy , the Lekki Massacre occurred on October 20th, 2020. One user named  @magnetvikki  describes the event as “souls [were] being wasted at Lekki Tollgate” and continues to mention another user named  @dj_switchaholic  who they claim had personally witnessed the events. Another user named  @KolawoleFalodun  characterizes the event as “soldiers shooting at unarmed #EndSARS protestors” without any justification or closure to the families of those who were killed. A heartbreaking question that many of the users appear to be asking: “who gave the orders?” remains unanswered. Several users, including  @ajibola005  and  @Dremanix , mention an account named  @jidesanwoolu , who was also identified by Voyant as a frequently used name. This username actually belongs to Babajide Sanwo-Olu who serves as the Governor of Lagos State in Nigeria. Another account named  @MBuhari  is mentioned, which belongs to Nigerian President Muhammmadu Buhari and also appeared on the list of key people identified by Voyant. But how are these two government officials involved in the #EndSARS movement or related to the Lekki Massacre?

Many users Tweeting about the Lekki Massacre and seeking justice for the actions of the Nigerian police turned to government officials like Muhammadu Buhari and Babajide Sanwo-Olu. A total of 338 Tweets mentioned Buhari's account, 136 mentioned Sanwo-Olu's, and 140 mentioned both users concurrently. I used the TopicModelingTool to analyze the contents of these Tweets and understand how they might be interconnected or possibly relate to the Lekki Massacre. This tool determined topics by evaluating the words contained in each grouping of Tweets and placing commonly used words together under a particular topic. It also provided a count of the total number of words from each group that were used to create their designated topic. This value helps evaluate how much a topic relates to each politician. Once each topic has been created, the tool provides a list of the top words that fall under each topic and the researcher is responsible for interpreting these words and coming up with a label for them. The following topics were generated from the Tweets that mentioned the Nigerian politicians:

Documenting the Now. (n.d.). DocNow. Retrieved March 6, 2021, from  https://www.docnow.io/ 

Yang, G. (2016). Narrative Agency in Hashtag Activism: The Case of #BlackLivesMatter. Media and Communication, 4(4), 13–17.  https://doi.org/10.17645/mac.v4i4.692