By Silas Jonathan and Kemi Busari
This study examines the degree of bot operation on Twitter in the build-up to the 2023 general elections in Nigeria. Botomoter and Botsight tools are deployed to extract and distinguish real accounts from accounts rated as bots. Findings revealed a considerable presence of bot accounts following the five candidates selected for the study ahead of the 2023 general elections. Most of the bot accounts were found to have been created in 2022, significantly in August and September. The study also found that bot accounts significantly interact with popular hashtags on the candidates and the election.
Over the last decade and a half, social media has increasingly gained importance in political communication in democracies. Individuals and political actors use social media platforms to debate topics or to execute online political campaigns (Amleshwaram, A. A., Reddy, N., Yadav, S., Gu, G., & Yang, C., 2013). Twitter, an interactive social media platform with millions of users worldwide and in Nigeria, has proven to be an effective platform that engineers political discourses and functions as a channel deployed by some influencers to alter users’ political preferences (Nguyen, 2018). A 2015 study by Demos, a UK-based market research firm, revealed that Twitter was ten times more active in Nigeria during the 2015 election period than at “normal” times. Over 12.4 million tweets were made about the elections over the period, and 1.38 million unique Twitter users posted content on the platform.
Although the use of Twitter for political purposes is a recent development in Nigeria, the tool is becoming increasingly popular among key political actors and stakeholders. For instance, in the 2015 elections, and even more recently in the 2019 elections, political parties, mainly the Peoples Democratic Party (PDP) and the then main opposition party, the All Progressives Congress (APC), and their key candidates used Twitter to market themselves and influence voter decision making. This increasing influence of social media, especially Twitter, in the heat of Nigerian electoral politics raises the question of how it is used to shape the political landscape in Nigeria. Social robots referred to as bots can be programmed to post information about the news of any kind automatically and even provide help during emergencies (Yan, Peter & Dennis 2021). Also, as amplifiers of messages such as hashtags, bots can be technological instruments ready to serve their creators (Keller, Spencer, Danet & Jan 2020).
Although Twitter bots are not real accounts, they help drive traffic to a user or a hashtag by operating like accounts controlled by human users by retweeting, liking posts, and following other accounts (Keller, F., Schoch, D., Stier, S., & Yang, J. H., 2019). Bots are unique and dynamic in their operations. (Murthy, Jane, Tom & Guise 2016) disclosed that Twitter bot accounts differ in their level of sophistication – from low-level accounts that merely aggregate information from websites to produce simple messages to a more sophisticated social bot that can be conversational and aim at posing as a human.
In January 2017, Twitter announced a response to concerns about how the platform was used during the 2016 US Presidential election to spread information disorder (Shevtsov, Konot, Kane & Simeon 2022). As a result, the company removed over 220,000 accounts responsible for over 2.2 billion low-quality tweets that violated Twitter’s rules. Twitter also confirmed that over 50,000 Russia-linked bots attempted to interfere with the 2016 US election (Shevtsov et al., 2022). In response to this challenge, the company escalated its battle against information disorder and hate speech by suspending over one million bots daily for months throughout 2018. Henriksen (2022) noted that this was a remarkable shift to lessen the flow of disinformation on the platform.
Nonetheless, (Papakyriakopoulos, Ahmad & Stephen 2022) found that bots are still active on Twitter, even anchoring key political-related hashtags. This finding was further reinforced by Schoenmueller (2022), who found that automated accounts are exceptionally efficient in spreading low-credibility content and amplifying the visibility of famous figures on Twitter. Likewise, (Beutel, Kata, Tom & Liam 2022) revealed how bot accounts on Twitter target influential people, bombarding them with hateful content and influencing public opinion during notable events. However, Benzie (2022) found that bots interact with genuine users based on their political affiliations to spread misinformation, implying that bots amplify tweets and are strategically set to execute an intentional task, a function that Sushi (2022) describes as a massive weapon of mass destruction that needs no gigantic arsenal.
While many studies have explored the operations of social bots in western polity, the situation in Africa, especially in Nigeria, seems to have been off the radar. A recent study by Kayode-Adedeji, T., & Nwakerendu, I. (2022) revealed a rising interest by various political parties to influence political conversations on social media ahead of the 2023 general elections in Nigeria. This finding speaks of an ongoing subtle and expansive operation targeted at the forthcoming presidential election. To this effect, the noticeably increasing followership of the leading political candidates on Twitter since late 2021 and the regular trend of politically affiliated hashtags on Twitter, such as #obidients, #Atikulate, and #BATtified not only sparks doubts but steers the need for an urgent study. Therefore, this study aims to determine the degree of bot operation around the 2023 presidential candidates on Twitter in the build-up to the elections.
2.0 Research methodology and conceptualisation
This study aims to determine the degree of bot operation on ‘Nigeria Twitter’ in the build-up to the 2023 general elections in Nigeria. Twitter is a favourable case study due to its relevance in Nigerian electoral politics, specifically in political campaigns and discussions. The analysis is based on bot accounts following the selected presidential candidates and bots’ activities (also known as astroturfing) around trending hashtags related to these candidates.
Since the official commencement of political activities in March 2022, some of these candidates’ followers have increased sporadically, and hashtags related to them have been trending. The study primarily focused on bot accounts following the presidential candidates. However, for the sake of data reliability and validity, the study limited the candidates to those who are:
1. Approved by the Independent National Electoral Commission (INEC);
2. Verified on the Twitter platform.
The above benchmarks are vital because they validate the data’s reliability and offer the assurance that the researchers were dealing with the actual account of the candidates. Of the 18 INEC-approved candidates, nine were not verified as of September 2022, when data was collected for this research, and four had no accounts. As such, the researchers were left with five verified candidates who met the selection criteria. The table below shows the five candidates verified on Twitter. Their official Twitter handles, and their total number of followers in September 2022.
Table 1. Names, usernames and followers of verified candidates
|S/N||Candidate||Political Party||Twitter Handle||No. of followers (September 2022)|
|Bola Ahmed Tinubu |
VP: Kashim Shettima
|All Progressives Alliances (APC)||Bola Ahmed Tinubu@officialABAT|
VP: Kashim Shettima @shettimaSM
|1.4 million, (1,432,102)|
VP: Ifeanyi Arthur Okowa
|Peoples Democratic Party (PDP)||Atiku Abubakar @atiku|
VP: Ifeanyi Arthur Okowa@IAOkowa
4.5 million, (4,569,132)
|Rabiu Musa Kwankwaso|
VP: Isaac Idahosa
|New Nigeria Peoples Party (NNPP)||Rabiu Musa Kwankwaso @kwankwasoRM|
VP: Isaac Idahosa@bishopiidahosa
|Peter Obi |
VP: Datti Baba-Ahmed
|Labour Party (LP)||Peter Obi@PeterObi|
VP: Datti Baba-Ahmed@dattibabaahmed
|2.1 million, (2,128,909)|
|Omoyele SoworeVP: Haruna Magashi||African Action Congress||@YeleSowore|
2.2 Botsight and Botometer
Having identified these accounts, Botsight and Botometer were used to evaluate and distinguish bot accounts from the real Twitter following of the five candidates. Botometer is an open-source tool developed at Indiana University to distinguish bot accounts from real ones. It does this by examining each Twitter account by aggregating elements of bot likelihood. The Botometer tool is accessed as a website (botometer.iuni.iu.edu). It rates any Twitter user’s followers with ratings ranging from zero to five, where zero is the most likely human account and five is the most likely bot.
Daniel Kats, a Senior Principal Researcher at Norton LifeLock, developed Botsight. The tool examines each Twitter account by constituting elements of bots. It functions as an extension on personal computers that rates Twitter accounts on the Twitter application. Botsight categorises its ratings into two, ranging from one to 99 per cent for both categories. The first category is for accounts rated as bots. Where the percentage is higher and is presented in red or orange, the account is “almost certainly a bot.” In contrast, accounts presented in yellow have “human and some bot-like” characteristics. The second category, on the other hand, is for real accounts. Here, a high percentage signals a real account. The results are presented in green, which signals an account is “almost certainly human,” and dark green indicating an account is “likely human.”
Both Botsight and Botometer perform similar tasks. However, the two tools are combined in this study for reliability and validity. Botsight was first used to rate the bots’ likelihood of followers on the Twitter platform, and 20 accounts rated as bots were then randomly picked from the results and further tested on Botometer. Both tools (with different rating systems) rated the selected 100 accounts (20 for each of the five candidates) as bots.
Hoaxy, a tool developed by Menczer, part of a group of researchers based in the US and China, was used to analyse the prevalence of bot accounts around some popular hashtags directly related to the five candidates. Hoaxy visualises two aspects of the spread of claims and topics relating to a specific theme: temporal trends and diffusion networks. Temporal trends (hashtags) plot the cumulative number of Twitter shares over time. The user can zoom in at any time interval. Diffusion networks display how claims spread from person to person. Each node is a Twitter account, and two nodes are connected if they pass a link to a story between those two accounts via retweets, replies, quotes, or mentions. Hoaxy also rates accounts likely to be bots and how they operate within the network. A five may indicate a large amount of bot automation, while a zero may indicate little to no automation. Bot scores are calculated using a machine learning algorithm trained to classify the level of automation an account presents.
Hoaxy uses colours to depict the activities around a particular hashtag (distinguished in colours). Therefore,
· Red: means the user is highly bot-like.
· Orange: user has both human and bot-like qualities.
· Yellow: the user is rated as likely human.
· Green: the user is rated as human.
· Blue: the user is rated as highly human.
2.4 Data gathering and analysis
Primarily, data for this study was collected using open-source tools. The followers of each candidate in this study were subjected to bots analysis using the Botsight tool. The Botometer tool was also used. The general principles of bot identification (accounts with numbers attached to their usernames, low or zero followers, few or no tweets, and recency in account creation) were further employed to verify the ratings from Botsight to ensure validity and reliability.
Each candidate’s bots following was totalled, categorised and exported to an Excel spreadsheet using the TwExport extension. The identified bot followers were analysed as independent entities to understand the dates they were created, the extent of their activities, and their followers, among others.
A list of hashtags was manually compiled based on their relevance to the 2023 elections and their direct relationship to the candidates. These hashtags were then analysed on the Hoaxy tool to track bot influence around their presence. The table below contains details of the hashtag and the candidate it relates to.
Table 2: lists the hashtags relevant to the 2023 general elections
|S/N||Hashtag/ Keyword||Candidate Associated With||Context|
|#obidients||Peter Obi||A hashtag coined from “Obi” by Peter Obi’s supporters. The term broadly means someone or persons supporting Peter Obi.|
|2.||#peterobi||Peter Obi||Hashtag support Peter Obi|
|3.||#BAT||Bola Ahmed Tinubu||Popular hashtag coined from Tinubu’s initials: Bola Ahmed Tinubu.|
|#Tinubu||Bola Ahmed Tinubu||A hashtag referenced to Tinubu.|
|7.||#Kwankwaso||Rabiu Musa Kwankwaso||Usually used to refer to Rabiu Musa Kwankwaso or his supporters.|
|8.||#RMK2023||Rabiu Musa Kwankwaso||The hashtag directly referenced Rabiu Musa Kwankwaso.|
|9.||#Atikulated||Abubakar Atiku||Mostly referred to the supporters or intending supporters of Abubakar Atiku.|
|10.||#Atiku||Atiku Abubakar||Hashtag related to Atiku Abubakar.|
|11.||#Sowore||Omoyele Sowore||Hashtag related to Omoyele Sowore.|
|12.||#Soworemagashi||Omoyele Sowore||Hashtag related to Sowore Sowore and his running mate, Haruna Magashi.|
Due to the researchers’ inability to access Twitter’s API, which would have offered broad and ready-made data about followers of each candidate in the study, the proportionate stratified sampling technique was utilised using a 10% fraction from the total number of bots following each of the candidates. (Rahman, Azad & Samrat 2022) assert that 10% is adequate for a sample size where a researcher can only access some population sizes. See details in Table 3 below. The data from the analyses were sorted, analysed and presented in charts, tables, and screenshots to draw out patterns.
Table 3: Bot followers per candidate and number of the extracted sample size
|S/N||Candidates||Number of Followers (August 2022)||Total No. of Bots||10% extract for Analysis|
|1.||Peter Obi||2.1 million, (2,128,909)||531, 000||53,100|
|2.||Atiku Abubakar||4.5 million, (4,569,132)||414,000||41, 400|
|3.||Rabiu Musa Kwankwaso||238,479||11,865||1,186|
|4.||Bola Ahmed Tinubu||1.4 million, (1,432,102)||248,000||24800|
2.5 Timeline for the study
The researchers did not set a specific timeline for collecting data for this research. However, data gathered shows that the latest bot following for any of the candidates dates back to November 2021 and the latest, September 2022. This puts the study period to be between November 2021 and September 2022.
2.6 Key concept
- Bots: An autonomous internet or social media program that can interact with systems or users.
- Twitter bots: A software program on Twitter that performs automated, repetitive, pre-defined tasks such as tweets, retweets, and likes.
- Independent National Electoral Commission (INEC): The legally mandated body instituted to conduct free and fair elections in Nigeria.
3.0 Key findings
Data for this research was gathered from Twitter using Botsight and Botometer to analyse the number of bot accounts following five presidential candidates. A total of 1.24 million bot accounts were found to be distributed across all five major candidates’ accounts from November 2021 to September 2022. Botometer and Botsight labelled all sampled accounts following the five candidates up to 100% bot rating; none of the accounts in our sample had ratings below 80%. An average bot rating of 95.7% was found, showing strong bot probability. Duplicate bot accounts were also found to follow each candidate to varying degrees. These accounts were identified and removed from the analysis.
Table 4: Duplicate bot accounts found and removed across all five candidates.
Removed: 220; total remaining: 123419
3.1 Bots following per candidate
The ratio of each candidate’s bots following was analysed. The analysis showed that while Atiku Abubakar had the highest number of followers on Twitter, at 4.5 million, Peter Obi’s account of 2.1 million followers had the highest percentage (25%) of the cumulative total bot accounts following the five candidates. This is suspected to stem from Peter Obi’s popularity over the past months. Obi is followed by Bola Tinubu (18%), Atiku Abubakar (9.2%), Rabiu Kwankwaso (5%) and Sowore (4%), respectively. The findings suggest that the higher number of bot followers a candidate has, the greater their popularity on Twitter.
3.2 Creation of bot accounts (by period)
Ninety-two per cent (92%) of bot accounts following the candidates were created in 2022 – from January 2022 to September 2022, with the highest number of accounts (93,672) created in September. Of the five candidates, Obi had the most bot account followers (47,134) created in 2022, mainly after the primary elections in June. This finding resonates with the spiral increase of his followership; Obi’s followers as of February 26, 2022, were 705,600 but increased by over 1.3 million to 2.1 million by September 2022.
3.3 Spiral increase in bot accounts created in 2022
The findings show a steady increase in bot accounts from January to September 2022. Significantly, Atiku, Obi and Tinubu cumulatively had a 65% increase in bot followers each month from July to August.
An almost 200% increase in bot account followership was observed in September, the month stipulated for the official commencement of political campaigns, with a sharp jump in the number of bot accounts created and following Atiku, Obi and Tinubu.
Additionally, the findings showed that Atiku, Obi and Tinubu had a minimum of 200 bot accounts following them each month. On the other hand, Sowore and Kwankwaso did not encounter their first 100 bot accounts’ following until April and August, respectively.
|Bot accounts creation from January – September 2022|
The findings suggest a deliberate strategy by faceless actors to contribute to and influence discussions inorganically through bot accounts at a crucial time in the campaign.
3.4 Bot accounts and their followers
The chart below (Figure 9) shows the followership and following of the bot accounts. The findings show that the accounts had very few followers and followed few accounts. On average, an account in the sampled data had a follower count of approximately five and followed an average of 78 accounts. The highest number of followers an account had was 58. Most of them had between zero and ten accounts following them, with all bot accounts following Kwankwaso having a zero follower count.
4.0 Hashtag Analysis
This section covers how real Twitter accounts and bots interacted with tweets containing hashtags or keywords directly related to a particular presidential candidate. For this analysis, data was gathered using Hoaxy and Botomoter. The analysis is limited to only September, the month that registered the highest bot creation. Two hashtags related to each candidate were selected based on how they trended in the month (see figure 2). In the results presented by Hoaxy, the red dashboard signals bots’ activities around a particular hashtag. The orange colour signals bots that have both human and machine qualities. Results from this analysis show bot activities around all the selected hashtags related to the candidates. Interestingly, the level of bot activities related to candidates resonated with their number of bot followers. As such, the bot following of a candidate was found to be proportionate to the number of bots around related hashtags.
4.1 Atiku Abubakar (PDP)
As of September 2 2022, Atiku had 4.5 million followers, of which 414,000, representing 9.2%, were bots. Many of his bot followers were created between August (1,482) and September (35,843) 2022. The analysis on the two hashtags directly related to the PDP candidate, #atikulated and #atiku, showed significant bot activity around them. Of the 1,003 accounts that interacted with #atikulated, 195 (19%) were rated as bots. At the same time, #Atiku had 230 (23%) accounts out of 962 bots interacting with the hashtag. The researchers speculate that more bot activity around these hashtags will increase with the intensity of the campaigns, especially as the election draws near.
4.2 Bola Ahmed Tinubu
A bot analysis revealed that 17.1% (248,000) of Tinubu’s 1.4 million (by September) were bots. Like Atiku, many bot accounts following Tinubu were created in August (3,559) and increased in September (17,257) 2022.
An analysis of the two hashtags related to the candidate, #Tinubu and #BAT, showed some bot activity around them. #Tinubu had 967 accounts pushing narratives around it, of which 236 (24%) were rated as bots. #BAT, on the other hand, had 718 accounts around it, and 219 (30%) were rated as bots.
Although this analysis is limited to September, the numbers will likely increase as the election draws near.
4.3 Peter Obi
A bot analysis of Peter Obi’s followers revealed that 531,000, representing 26.55%, were bots. This data corresponds with the previous findings, where Obi was found to have 39,782 bots followers, the highest bot count in September. Analysing #Obidient and #peterobi hashtags revealed bot activity around them. These bot accounts influencing the hashtags are spread across the platform and showed that #Obidients had 201 (21%) out of 917 accounts rated bots. In contrast, #peterobi had 165 (18%) out of 917 accounts rated as bots.
This finding indicates a growing bot operation around topics and hashtags related to Peter Obi. Although this analysis is carried out in September, it will likely increase in the months before the 2023 general elections.
4.4 Rabiu Musa Kwankwaso
Hashtags related to Kwankwaso, #RMK2023 and #kwankwaso also had some bot activities around them. However, unlike hashtags related to Atiku and Obi, bots’ activities around Kwankwaso’s hashtags were not extensive. The #RMK2023 had 639 accounts operating around it, and 67 (10%) were rated as bots. In comparison, #kwankwaso had 132 (17%) out of 750 accounts rated as bots.
4.5 Omoyele Sowore
Out of 789,215 accounts following Sowore (as of September 26), almost 10% (31,529) were bot accounts created in August and September 2022.
An analysis of hashtags #sowore and #soworemagashi revealed some levels of bot activities. The #soworemagashi2023 had 71 (14%) out of 485 accounts rated as bots, while #sowore was found to have 190 (22%) accounts out of 922 rated as bots. Although less extensive than that of Atiku, Obi or Tinubu, these bots operating around these hashtags show a deliberate effort to promote conversations around Sowore.
5.0 Discussion of findings
The study found a massive bot presence in the Nigerian Twitter space related to the 2023 general elections. Out of the 9.8 million accounts following all five major candidates on Twitter, 1.24 million (12.7%) were bots. About 80% of the bot accounts were created in 2022, mainly within two months before the official commencement of campaigns. While Atiku Abubakar had the highest followers on Twitter, Peter Obi had more bot account followers than any other candidates. Interestingly, the study also found that these bot accounts sprouted massively after all the political parties had selected a candidate in the primary elections conducted in June 2022. The study also established that all the candidates had a massive increase in followers on Twitter within just nine months, that is, January to September 2022
An analysis of the bot accounts showed that they had few followers and very few activities, as most of them neither tweeted nor retweeted. This aligns with Yamaguchi (2010), who explained that some bot accounts rarely post messages but follow many user accounts to get information from other users’ tweets.
Additionally, we noticed a relative bots activity around the popular hashtags related to the candidates. All ten selected hashtags had a significant bot activity around them. Although hashtags around Atiku had higher bot activities than all the other candidates, we speculate they will all catch up with all the candidates and increase as the months to the elections draws closer.
6.0 Limitations to the study
A major limitation of this study was the restriction of data to only five out of the 18 presidential candidates. The limitation was necessary to collect reliable and valid data from the accounts of the candidates that are verified on Twitter. Another major limitation was the inability to access Twitter API to collect data. This forced the study to draw a sample size from the collated data.
The study identified the presence of bots in the Nigerian political Twitter space. The massive number of bot accounts created within a few months shows a deliberate effort by some political actors to influence political conversations on social media or alter voters’ choices. While we cannot immediately pinpoint the specific reason for creating these bot accounts, the fundamental aim of their operations cannot be farfetched. As such, it will also be crucially helpful to identify the faceless actors perpetuating this scheme and know if these candidates are aware of it.
Alothali, E., Zaki, N., Mohamed, E. A., & Alashwal, H. (2018). Detecting social bots on Twitter: A literature review. 2018 International Conference on Innovations in Information Technology (IIT). https://doi.org/10.1109/innovations.2018.8605995
Amleshwaram, A. A., Reddy, N., Yadav, S., Gu, G., & Yang, C. (2013). Cats: Characterizing automation of twitter spammers. 2013 Fifth International Conference on Communication Systems and Networks (COMSNETS). https://doi.org/10.1109/comsnets.2013.6465541
Benzie, A. and R. Montasari (2022). Artificial Intelligence and the Spread of Mis-and Disinformation. Artificial Intelligence and National Security, Springer: 1-18.
Beutel, I., Kirschler, O., & Kokott, S. (2022). How do fake news and hate speech affect political discussion and target persons and how can they be detected? Central and Eastern European EDem and EGov Days, 342, 37–81. https://doi.org/10.24989/ocg.v.342.2
Caldarelli, G., De Nicola, R., Del Vigna, F., Petrocchi, M., & Saracco, F. (2020). The role of bot squads in the political propaganda on Twitter. Communications Physics, 3(1). https://doi.org/10.1038/s42005-020-0340-4
Chang, H.-C. H., Chen, E., Zhang, M., Muric, G., & Ferrara, E. (2021). Social Bots and social media manipulation in 2020. Handbook of Computational Social Science, Volume 1, 304–323. https://doi.org/10.4324/9781003024583-21
Hajli, N., Saeed, U., Tajvidi, M., & Shirazi, F. (2021). Social Bots and the spread of disinformation in social media: The Challenges of Artificial Intelligence. British Journal of Management, 33(3), 1238–1253. https://doi.org/10.1111/1467-8551.12554
Henriksen, L., & Wang, C. (2022). Hello, Twitter bot!: Towards a bot ethics of response and responsibility. Catalyst: Feminism, Theory, Technoscience, 8(1). https://doi.org/10.28968/cftt.v8i1.36203
Kayode-Adedeji, T., & Nwakerendu, I. (2022). The Dissemination of fake news on social media: A demographic analysis of audience involvement. In European Conference on Social Media (Vol. 9, No. 1, pp. 289-297).
Keller, F., Schoch, D., Stier, S., & Yang, J. H. (2019). Political astroturfing on Twitter: How to coordinate a disinformation campaign. https://doi.org/10.31235/osf.io/a5gk6
Keller, T. R. and U. Klinger (2019). “Social bots in election campaigns: Theoretical, empirical, and methodological implications.” Political Communication 36(1): 171-189.
Kenny, R. J. (2022). Why Some Are Better than Others at Detecting Social Bots: Comparing Baseline Performance to Performance with Aids and Training, Carnegie Mellon University.
Macey-Dare, R. (2022). “Elon’s Unhappy Twitter Meal-Big MAC+ Spam with Reduced mDAU.” Available at SSRN 4192705.
Murthy, D. and Sean, A. (2016). “Automation, algorithms, and politics| Bots and political influence: A sociotechnical investigation of social network capital.” International journal of communication 10: 20.
Ndlela, M. Ilhem A., and Abeer A., (2020). Social media algorithms, bots and elections in Africa. Social Media and Elections in Africa, Volume 1, Springer: 13-37.
Nguyen, J. (2018). “Politics and the Twitter revolution: A brief literature review and implications for future research.” Social Networking 7(4): 243-251.
Papakyriakopoulos, O. and E. Goodman (2022). The Impact of Twitter Labels on Misinformation Spread and User Engagement: Lessons from Trump’s Election Tweets. Proceedings of the ACM Web Conference 2022.
Parmelee, J. H. and S. L. Bichard (2011). Politics and the Twitter revolution: How tweets influence the relationship between political leaders and the public, Lexington books.
Pastor-Galindo, J.,Meryl A., and Adriana A., . (2022). “Profiling users and bots in Twitter through social media analysis.” Information Sciences.
Rahman, M. and Salma C. (2022). “Sampling techniques (probability) for quantitative social science researchers: a conceptual guidelines with examples.” Seeu Review 17(1): 42-51.
Schoenmueller, V., et al. (2022). “Who will share Fake-News on Twitter? Psycholinguistic cues in online post histories discriminate Between actors in the misinformation ecosystem.” arXiv preprint arXiv:2203.10560.
Shahi, G. K. and K. Tsoplefack (2022). Mitigating Harmful Content on Social Media Using An Interactive User Interface. International Conference on Social Informatics. Springer.
Shevtsov, A. Konot,S., Kane D., Suarez N., & Simeon P. (2022). Identification of Twitter Bots Based on an Explainable Machine Learning Framework: The US 2020 Elections Case Study. Proceedings of the International AAAI Conference on Web and Social Media.
Smith, J. (2022). DIGITAL FORENSICS: 17 TOOLS TO DISCERN FACT FROM FICTION ONLINE. EDULEARN22 Proceedings, IATED.
Soedarsono, D. and Karr, J (2020). “Political leaders and followers’ attitudes: Twitter as a tool for political communication.” Journal of Adverstisment Research in Dynamical y Control Systems 12(6): 1615-1626.
Suárez-Serrato, P., Baur, F, and Soph. H (2016). On the influence of social bots in online protests. International Conference on Social Informatics, Springer.
Varol, O. (2022). “Should we agree to disagree about Twitter’s bot problem?” arXiv preprint arXiv:2209.10006.
Wimmer, R. and Dominick, J. (2013). Mass media research, Cengage learning.
Woolley, S. C. (2016). “Automating power: Social bot interference in global politics.” First Monday.
Yamaguchi, Y., Meryl, A., Adriana, A., (2010). Turank: Twitter user ranking based on user-tweet graph analysis. International Conference on Web Information Systems Engineering, Springer.
Yan, H. Y., Peter M., & Dennis, B.S (2021). “Asymmetrical perceptions of partisan political bots.” New Media & Society 23(10): 3016-3037.