Academia.eduAcademia.edu

Figure 3 – uploaded by Philipe Freitas Melo

See full PDFdownloadDownload figure
Table 3: Topics extracted from the English tweets that include WhatsApp, Telegram, and Discord group URLs.

Table 3 Topics extracted from the English tweets that include WhatsApp, Telegram, and Discord group URLs.

Related Figures (15)

Table 1: An overview of the characteristics of the different messaging platforms, highlighting some of the differences.
Table 1: An overview of the characteristics of the different messaging platforms, highlighting some of the differences.
Table 2: Overview of our datasets.
Table 2: Overview of our datasets.
Figure 1: Number of group URLs discovered on each day during our Twitter data collection, showing that Twitter is a rich data source for discovering public groups on the different messaging platforms. Demystifying the Messaging Platforms’ Ecosystem Through the Lens of Twitter
Figure 1: Number of group URLs discovered on each day during our Twitter data collection, showing that Twitter is a rich data source for discovering public groups on the different messaging platforms. Demystifying the Messaging Platforms’ Ecosystem Through the Lens of Twitter
Figure 2: CDF of number of tweets for each group URL over the entire dataset. A large percentage of the group URLs are shared only once. We begin our analyses with the number of group URLs discov- ered on Twitter for the three messaging platforms (see Fig. 1). We report three different metrics: (1) all group URLs discovered on Twitter; (2) the number of unique group URLs per day; and (3) the number of new group URLs per day (ie., excluding group URLs al- ready observed on previous days). WhatsApp appears, per Fig. 1, to be the most “private” messaging platform: We discover fewer group URLs belonging to WhatsApp than that of Telegram and Discord, despite WhatsApp being a much larger and widely used messaging
Figure 2: CDF of number of tweets for each group URL over the entire dataset. A large percentage of the group URLs are shared only once. We begin our analyses with the number of group URLs discov- ered on Twitter for the three messaging platforms (see Fig. 1). We report three different metrics: (1) all group URLs discovered on Twitter; (2) the number of unique group URLs per day; and (3) the number of new group URLs per day (ie., excluding group URLs al- ready observed on previous days). WhatsApp appears, per Fig. 1, to be the most “private” messaging platform: We discover fewer group URLs belonging to WhatsApp than that of Telegram and Discord, despite WhatsApp being a much larger and widely used messaging
Figure 3: Percentage of tweets that contain hashtags/mentions and percentage of tweets that are retweets. Twitter users tend to not use hashtags and use mentions when sharing WhatsApp, Telegram, and Discord groups. Tweets with Telegram groups are more likely to get retweeted compared to the other platforms.
Figure 3: Percentage of tweets that contain hashtags/mentions and percentage of tweets that are retweets. Twitter users tend to not use hashtags and use mentions when sharing WhatsApp, Telegram, and Discord groups. Tweets with Telegram groups are more likely to get retweeted compared to the other platforms.
Figure 5: Staleness: time difference between the appearance of the group on Twitter and its creation date. Older Telegram and Discord groups are shared on Twitter, while shared WhatsApp groups are “fresh”.
Figure 5: Staleness: time difference between the appearance of the group on Twitter and its creation date. Older Telegram and Discord groups are shared on Twitter, while shared WhatsApp groups are “fresh”.
Figure 6: Analysis on how long groups are accessible un- til they get revoked. A substantial percentage of groups are not accessible during our first observation (especially Discord groups). Demystifying the Messaging Platforms’ Ecosystem Through the Lens of Twitter
Figure 6: Analysis on how long groups are accessible un- til they get revoked. A substantial percentage of groups are not accessible during our first observation (especially Discord groups). Demystifying the Messaging Platforms’ Ecosystem Through the Lens of Twitter
Figure 7: Distributions of number of members per group, percentage of online members over all members, and group size change over time (between first and last observation).
Figure 7: Distributions of number of members per group, percentage of online members over all members, and group size change over time (between first and last observation).
Figure 8: Percentage of messages in each message type. Text is the predominant type across all messaging platforms.
Figure 8: Percentage of messages in each message type. Text is the predominant type across all messaging platforms.
Figure 9: Distribution of the number of messages shared in the groups of each platform and their users.
Figure 9: Distribution of the number of messages shared in the groups of each platform and their users.
(a) Mean number of messages for each group per day
(a) Mean number of messages for each group per day
Table 4: Statistics on users’ sensitive PII that are exposed from the three messaging platforms. Table 5: Number and percentage of Discord users whose their accounts on other platforms are exposed.
Table 4: Statistics on users’ sensitive PII that are exposed from the three messaging platforms. Table 5: Number and percentage of Discord users whose their accounts on other platforms are exposed.
6 PRIVACY IMPLICATIONS
6 PRIVACY IMPLICATIONS
Related topics:
SociologyComputer ScienceSocial Network Analysis (SNA)Mobile social networkingInformation Security and PrivacyPrivacy and data protectionInternet privacyMobile Social Networking
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved