Abstract
Social media sites such as Facebook and Twitter provide highly granular time-stamped data about the interactions and communications between people and provide us unprecedented opportunities for empirically testing theory about information flow in social networks. Using publicly available data from Twitter’s free API (Application Program Interface), we track the adoption of popular hashtags in Nigeria during 2014. These hashtags reference online marketing campaigns, major news stories, and events and issues specific to Nigeria, including reactions to the kidnapping of 276 schoolgirls in Northeastern Nigeria by the Islamic extremist group Boko Haram. We find that hashtags related to Nigerian sociopolitical issues, including the #bringbackourgirls hashtag, which was associated with protests against the Nigerian government’s response to the kidnapping, are more likely to be adopted among densely connected users with multiple network neighbors who have also adopted the hashtag, compared to mainstream news hashtags. This association between adoption threshold and local network structure is consistent with theory about the spread of complex contagions, a type of social contagion which requires social reinforcement from multiple adopting neighbors. Theory also predicts the need for a critical mass of adopters before the contagion can become viral. We illustrate this with the #bringbackourgirls hashtag by identifying the point at which the local social movement transforms into a more widespread phenomenon. We also show that these results are robust across both the follow and reply/mention/retweet networks on Twitter. Our analysis involves data mining records of hashtag adoption and of the social connections between adopters.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For simplicity, Centola and Macy used an absolute threshold expressed in terms of the absolute number of adopting neighbors, but their results hold also for relative thresholds expressed in terms of the fraction of adopting neighbors.
The collection of data from Nigeria was related to a previously funded effort, unrelated to this work.
It should be noted that spammers may also use popular hashtags to gain attention for their tweets. Since we mostly focus on the early part of a tag’s lifespan, we assume that this effect is negligible. Also, there are likely many tweets about the topic of the hashtag that do not contain the hashtag. While some data will be missed in ignoring these tweets, the hashtag convention is so pervasive on Twitter that we assume the lower recall resulting from our retrieval strategy will not alter the results drastically.
Note that these fields contain the same information, but with reversed ordering of the coordinates.
Only 1 % of users had geotags in their profile strings and are excluded from this discussion.
In the case where the total days examined D is less than 10, we use the first D days.
News stories and dates were obtained via a Lexis/Nexis search.
References
Barash V (2011) The dynamics of social contagion. PhD thesis, Cornell University
Barash V, Cameron C, Macy M (2012) Critical phenomena in complex contagions. Soc Netw 34(4):451–461
BBC (2014) #BBCtrending: The man who 'disappeared' in Nigeria - #FreeCiaxon. Retrieved from http://www.bbc.com/news/blogs-trending-26979508. Accessed 23 Dec 2015
Centola D, Macy M (2007) Complex contagions and the weakness of long ties1. Am J Sociol 113(3):702–734
Geonames (2015) Retrieved from http://www.geonames.org. Accessed 23 Dec 2015
Granovetter M (1978) Threshold models of collective behavior. Am J Sociol 8(6):1420–1443
International Telecommunications Union (2015) Statistics: percentage of individuals using the internet. Retrieved from https://www.itu.int/en/ITU-D/Statistics/Pages/stat/default.aspx. Accessed 23 Dec 2015
Know Your Meme (2015) #WeAreAllMonkeys. Retrieved from http://knowyourmeme.com/memes/events/weareallmonkeys. Accessed 23 Dec 2015
Leskovec J, Backstrom L, Kleinberg J (2009) Meme-tracking and the dynamics of the news cycle. In: Proceedings of the 15th ACM SIGKDD international conference on knowledge discovery and data mining, pp 497–506. ACM
Marwell G, Oliver PE, Prahl R (1988) Social networks and collective action: a theory of the critical mass. iii. Am J Sociol 502–534
Milgram S (1967) The small world problem. Psychol Today 2(1):60–67
Nahon K, Hemsley J (2013) Going viral. Polity Press, Cambridge
Odufuwa F (2012) Understanding what is happening in ICT in Nigeria. Technical report, Research ICT Africa
Romero DM, Meeder B, Kleinberg J (2011) Differences in the mechanics of information diffusion across topics: idioms, political hashtags, and complex contagion on twitter. In: Proceedings of the 20th international conference on world wide web, pp 695–704. ACM
Watts DJ, Strogatz SH (1998) Collective dynamics of ‘small-world’ networks. Nature 393(6684):440–442
Weng L, Menczer F, Ahn YY (2013) Virality prediction and community structure in social networks. Scientific reports, 3
Wikipedia (2014) YesAllWomen. Retrieved from https://en.wikipedia.org/wiki/YesAllWomen. Accessed 23 Dec 2015
Wu S, Tan C, Kleinberg J, Macy M (2011) Does bad news go away faster? In: Fifth international AAAI conference on weblogs and social media ICWSM. AAAI
Acknowledgments
This work was funded by the United States Air Force Office of Scientific Research (AFOSR) through the Minerva Initiative under Grant FA9550-15-1-0036.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Fink, C., Schmidt, A., Barash, V. et al. Complex contagions and the diffusion of popular Twitter hashtags in Nigeria. Soc. Netw. Anal. Min. 6, 1 (2016). https://doi.org/10.1007/s13278-015-0311-z
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s13278-015-0311-z