UploadLanguage (EN)Support
BusinessMobileSocial MediaMarketingTechnologyArt & PhotosCareerDesignEducationPresentations & Public SpeakingGovernment & NonprofitHealthcareInternetLawLeadership & ManagementAutomotiveEngineeringSoftwareRecruiting & HRRetailSalesServicesScienceSmall Business & EntrepreneurshipFoodEnvironmentEconomy & FinanceData & AnalyticsInvestor RelationsSportsSpiritualNews & PoliticsTravelSelf ImprovementReal EstateEntertainment & HumorHealth & MedicineDevices & HardwareLifestyle
Upload
Sign in
Christoph Lutz, profile picture
Uploaded byChristoph Lutz
915 views

The WorldBot 2014

The document discusses the analysis of Twitter activity surrounding the 2014 World Cup, focusing on the categorization of bots and their behavior patterns. It outlines methods for bot detection and reveals a diverse range of bots, identifying categories labeled as #good, #bad, and #ugly, with the latter being particularly challenging to classify. The study concludes that World Cup tweeting is not heavily bot-driven and highlights the complexity of distinguishing bot behavior.

Data & Analytics◦Sports◦Technology◦
Related topics:
Data Visualization Techniques•
Download to read offline
1 / 34
2 / 34
3 / 34
4 / 34
5 / 34
6 / 34
7 / 34
8 / 34
9 / 34
10 / 34
11 / 34
12 / 34
13 / 34
14 / 34
15 / 34
16 / 34
17 / 34
18 / 34
19 / 34
20 / 34
The #WorldBot 2014 The #Good, the #Bad and the #Ugly Tommaso Elli ~ tommaso.elli@gmail.com Carlo De Gaetano ~ carlo.dgtn@gmail.com Christoph Lutz ~ christoph.lutz@unisg.ch Myrthe Bil ~ myrthebil@gmail.com Iris Beerepoot ~ imbeerepoot@gmail.com Claudio Coletta ~ claudio.coletta@gmail.com
Introduction ➔ WorldCup 2014: event with huge Twitter activity ➔ Much of the Twitter stream is structured along #hashtags ➔ Bots are represented with interesting possibilities of outreach and attention
Research Questions ➔ Can we categorize the bots surrounding the World Cup? ➔ Can we identify any behaviour pattern in the accounts we identify as bots? ➔ (Do different games’ nationality attract bots in different ways?)
Methods of Data Collection ➔ 1% of Twitter sample in TCAT ➔ All teams represented at least once
Methods for Bot Detection ➔ User account name and user description analysis ➔ Device/source analysis ➔ FF ratio (Zhu et al., 2012) ➔ User Co-Hashtag Analysis ➔ Close Reading
Results
Interactive Viz
Bot Detection (pt. 1) ➔ User account name and user description detected bots ➔ Sample of ~2500 supposed users
Bot Detection (pt. 1)
Bot Detection (pt. 1)
Bot Detection (pt. 1) ➔ If you’re out of the mass, you deserve some attention ➔ If you have thousands of friends/followers, but just few tweets, you’re probably a bot ➔ It’s interesting to analyze the friends/followers ratio
Bot Detection (pt. 2) ➔ Users who declare to be a bot ➔ X Axis: Friends/Followers Proportion ➔ Y Axis: Total number of used #tags
Bot Detection (pt. 2) ➔ We consider declared bots as good bots ➔ Their metrics put them in the left-bottom corner ➔ Is that the good bots corner?
The #Good
Detecting the #Bads? Users in these clusters have normal FF ratios and don’t carry special bot attributes in their bios or usernames… still they are bots.
The #Bad
➔ The “ugly” works as a residual category, often half “human” half “non-human” (Cyborg) ➔ They engage followers through specific hashtags (i.e. “potato”) and tend to have more followers than friends ➔ Their content involves the emotional sphere ➔ Seems very promising for further qualitative analysis Detecting the #Ugly?
- Une_Patate - iinokinji - iAnnaBottle - Tiret_Tiret The #Ugly RT @orsolawa: Oh c'est chaud patate dès le début !!! #ESPCHI @
The Pope Could Be a Bot (A Bad One)
Bot Detection (pt. 2)
Bot Detection (pt. 2) ➔ Maybe yes, we have a good corner. ➔ Let’s run some bigger analysis, within our ~2500 sample!
Conclusion (and have a happy Comic Sans Day!) ❏ World cup tweeting is not very botted ❏ World cup bots are very diverse, it’s not all #good and #bad ❏ Large #ugly category in the middle that’s hard to categorize and make sense of ❏ Boundaries between #ugly and #good are blurry and more difficult to draw than between #good and #bad