Cross-Network Social Multimedia Computing - User-centric Social Multimedia Computing - page 85

Table 5.1 The examined topic list and “T-W” ratio
Topic : “T-W ratio”

Topic : “T-W ratio”

1 US presidential election 2012 : 2.08 11 Christmas 2012 : 1.93

2 Gangnam style : 2.20 12 Nexus 4 release : 4.56

3 Super Bowl 2012 : 2.42 13 iPhone 5 release : 2.57

4 Olympic 2012 : 2.29 14 Black ops II : 3.20

5 Justin Bieber : 1.79 15 Doctor who TV : 1.53

6 Star Wars : 1.07 16 Prometheus : 1.50

7 Dark knight rises : 1.61 17 Google glasses : 2.06

8 Minecraft game : 1.76 18 Call me maybe : 1.09

9 Galaxy S3 : 2.69 19 Spider Man : 1.41

10 Michael Jackson : 1.82 20 Whitney Houston : 2.98

Fig. 5.2 The global attention evolution to “Super Bowl 2012”

users on Twitter and YouTube respectively, and select 20 topics that have remarkable
involved users in both networks (as listed in Table 5.1 ). Figure 5.2 illustrates the
evolution of global attention to the topic “Super Bowl 2012”, i.e., the daily changing
number of involved users on Twitter and YouTube from Jan.1, 2012. It is shown that
there exist more active social activities on Twitter than on YouTube. By examining
the attention peaks, YouTube, and Twitter response to two and four of the subtopics
during the Super Bowl 2012 event, respectively. Around the subtopic “Division
Round” and “Super Bowl XLVII”, the attention on Twitter rises up earlier than that
on YouTube. This validates that Twitter is faster than YouTube in information sharing
and spread at the macro level.
At the micro level, to examine the temporal characteristics of overlapped users
between Twitter and YouTube, we represent each topic with frequent keywords in
respective networks and identify the beginning time when users get involved in the
topic. By aggregating the overlapped users who first get involved in the topic in
Twitter ( Twitter_first ) and in YouTube ( YouTube_first ), we calculate “T-W ratio”
for each topic by dividing the two user numbers (# Twitter_first /# YouTube_first ).
Table 5.1 shows the “T-W ratio” for the examined topics. Note that “T-W ratio”