The tweets they are a-changin': Evolution of twitter users and behavior
Open accessAbstract
The microblogging site Twitter is now one of the most popular Web destinations. Due to the relative ease of data access, there has been significant research based on Twitter data, ranging from measuring the spread of ideas through society to predicting the behavior of real-world phenomena such as the stock market. Unfortunately, relatively little work has studied the changes in the Twitter ecosystem itself; most research that uses Twitter data is typically based on a small time-window of data, generally ranging from a few weeks to a few months. Twitter is known to have evolved significantly since its founding, and it remains unclear whether prior results still hold, and whether the (often implicit) assumptions of proposed systems are still valid. In this paper, we take a first step towards answering these question by focusing on the evolution of Twitter's users and their behavior. Using a set of over 37 billion tweets spanning over seven years, we quantify how the users, their behavior, and the site as a whole have evolved. We observe and quantify a number of trends including the spread of Twitter across the globe, the rise of spam and malicious behavior, the rapid adoption of tweeting conventions, and the shift from desktop to mobile usage. Our results can be used to interpret and calibrate previous Twitter work, as well as to make future projections of the site as a whole.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Studies on Facebook and Twitter have found evidence of variations over time.
“Studies on both Facebook (Lampe et al., 2008) and Twitter (Liu et al., 2014) have found evidence of such variations.”
From Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries (2019) · cites this paper as Liu et al., 2014
There are several mechanisms rendering a message unavailable later on: explicit deletion by user, account switched to protected/private, account suspended by platform, or account deactivated.
“There are several mechanisms rendering a message unavailable later on (Liu et al., 2014): it was explicitly deleted by the user; the user switched their account to “protected” or private; the user's account was suspended by the platform; or the user deactivated their entire account.”
From Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries (2019) · cites this paper as Liu et al., 2014
Social data decays with time as users delete their content and accounts, and platform API terms of service prevent sharing of datasets as collected.
“Social data decays with time as users delete their content and accounts (Liu et al., 2014; Gillespie, 2015), and platforms APIs' terms of service prevent sharing of datasets as they are collected.”
From Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries (2019) · cites this paper as Liu et al., 2014
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Measurement and analysis of online social networks
Mislove, A., Marcon, M., Gummadi, K. P., Druschel, P., & Bhattacharjee, B. (2007) · Proceedings of the 7th ACM SIGCOMM conference on Internet measurement · doi:10.1145/1298306.1298311 · Cited by 3,057 · see citing works
Self-citationMeasuring user influence in twitter: The million follower fallacy
Cha, M., Haddadi, H., Benevenuto, F., & Gummadi, K. (2010) · Proceedings of the International AAAI Conference on Web and Social Media · doi:10.1609/icwsm.v4i1.14033 · Cited by 3,022 · cited 4× · see citing works
“, most users have many more friends that followers) indicating that Twitter is disassortative; similar observations have been made about Twitter (Cha et al. 2010) and other social networks (Mislove et al.”
BackgroundMethodInfluential” How old do you think I am?” A study of language and age in Twitter
Nguyen, D., Gravel, R., Trieschnigg, D., & Meder, T. (2021) · Proceedings of the International AAAI Conference on Web and Social Media · doi:10.1609/icwsm.v7i1.14381 · Cited by 287 · see citing works
Flagged first, then by citations
Research works printed but not linked: 7
- Almuhimedi, H.; Wilson, S.; Liu, B.; Sadeh, N.; and Acquisti, A. 2013. Tweets are forever: a large-scale quantitative analysis of deleted tweets. In CSCW.
- Bergsma, S.; Dredze, M.; Durme, B. V.; Wilson, T.; and Yarowsky, D. 2013. Broadly Improving User Classification via Communication-Based Name and Location Clustering on Twitter. In NAACL.
- Cheng, Z.; Caverlee, J.; and Lee, K. 2010. You Are Where You Tweet: A Content-based Approach to Geo-locating Twitter Users. In CIKM.
- Hecht, B.; Hong, L.; Suh, B.; and Chi, E. H. 2011. Tweets from Justin Bieber's heart: the dynamics of the location field in user pro- files. In CHI.
- Pennacchiotti, M., and Popescu, A.-M. 2011b. Democrats, repub- licans and starbucks afficionados: user classification in twitter. In KDD.
- Petrovic, S.; Osborne, M.; and Lavrenko, V. 2013. I Wish I Didn't Say That! Analyzing and Predicting Deleted Messages in Twitter. CoRR.
- Thomas, K.; McCoy, D.; Grier, C.; Kolcz, A.; and Paxson, V. 2013. Trafficking Fraudulent Accounts: The Role of the Underground Market in Twitter Spam and Abuse. In USENIX Security. 2012. Twitter Terms of Service. https://twitter.com/tos/.
Cited by
Sorted by citations
Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries
Olteanu, A., Castillo, C., Diaz, F., & Kıcıman, E. (2019) · Frontiers in Big Data · doi:10.3389/fdata.2019.00013 · Cited by 763 · see citing works
“Further, in the case of Twitter, much research relies on APIs that give access to at most 1% of the data (González-Bailón et al., 2014a; Joseph et al., 2014; Morstatter et al., 2014).”
BackgroundInfluentialHow the news media activate public expression and influence national agendas
King, G., Schneer, B., & White, A. (2017) · Science · doi:10.1126/science.aao1100 · Cited by 333 · see citing works
Accelerating dynamics of collective attention
Lorenz-Spreen, P., Mønsted, B. M., Hövel, P., & Lehmann, S. (2019) · Nature Communications · doi:10.1038/s41467-019-09311-w · Cited by 304 · see citing works
Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries
Olteanu, A., Castillo, C., Diaz, F., & Kiciman, E. (2016) · doi:10.2139/ssrn.2886526 · Cited by 763 · see citing works
- Authors
- Liu, Y., Kliman-Silver, C., & Mislove, A.
- Affiliations
- Northeastern University (US)
- Brown University (US)
- Venue
- Proceedings of the International AAAI Conference on Web and Social Media
- Published
- 2014-05-16
- Volume
- 8
- Issue
- 1
- Pages
- 305-314
- ISSN
- 2162-3449 (ISSN-L)
- 2334-0770
- OpenAlex
- W2185208392
- Cited by
- 126
- References
- 3
- Citations by year
- 2026: 1 · 2025: 1 · 2024: 3 · 2023: 6 · 2022: 4 · 2021: 14 · 2020: 10 · 2019: 11 · 2018: 18 · 2017: 12 · 2016: 25 · 2015: 15 · 2014: 6
- FWCI
- 17.80
- Type
- conference paper
- Language
- en
- Open access
- diamond
- doi.org
- Locations
- Proceedings of the International AAAI Conference on Web and Social Media (publishedVersion)
- AAAI Publications (The Association for the Advancement of Artificial Intelligence (AAAI)) (submittedVersion)
- Funders
- National Science Foundation
- Topics
- Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Complex Network Analysis Techniques
- Social Sciences › Communication › Social Media and Politics
- Physical Sciences › Computer Science › Computer Networks and Communications › Media and Digital Communication
- Keywords
- Twitter, microblogging
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
