what-when-how
In Depth Tutorials and Information
and thus need less formality in their communication) are more likely to be similar
to each other. hey also found that the more time a user spends per message, the
more likely it is that he or she is talking to someone of the opposite gender.
As already introduced, Reference 23 investigated the relation between social
networks and other outside information networks (that is, search engine in the
reference). Some other investigators have also studied this situation. Bhagat et
al. [21] used the blogs as a starting point to pull in data about other multiple
information networks and studied how these multiple networks interact with
each other. hey exploited three types of cross-information networks, which are
blog-blog (two different blog sites), blog-web, and blog-messaging networks.
Some interesting results were concluded, such as bloggers using the same blog-
ging service cite each other significantly more than those using other services
in the blog-blog situation, and the percentage of users that share any IM con-
tact decreases with their age. Adamic and Glance [24] analyzed the posts of
40 “A-list” blogs over a period of 2 months preceding the U.S. Presidential
Election of 2004 across multiple blog sites to study how often they referred
to one another and to quantify the overlap in the topics they discussed, both
within the liberal and conservative communities, and also across communities.
he results show that liberals and conservatives link primarily within their
separate communities, with far fewer crosslinks exchanged between them.
An interesting pattern that emerged was that conservative bloggers were more
likely to link to other blogs: primarily other conservative blogs, but also some
liberal ones.
3.2.2 Analysis of User Behaviors
hroughout the information propagation process in online social networks, user
behaviors [42] could be divided into two types: reading and posting. Reading behav-
ior means that users read the articles, pictures, or videos they are interested in. his
type of behavior is the prior step of users' posting behaviors, which are posting com-
ments and articles about what they have read. However, studying reading behaviors
is quite difficult because most SNS did not support the data about who had read
the article or picture. We find that just one research team has investigated it based
on the SNS's log data [31, 38]. he studying on posting behaviors is more popular.
Investigators have analyzed the behavior based on the social network structure,
users' content interests, and time factors [32, 33, 36, 37, 44, 47]. Note that there is
a significant relative research focus: mining the users' interests [34, 35, 50], which is
obviously helpful to analyzing and modeling posting and reading behaviors.
3.2.2.1 Reading Behavior
Furukawa et al. [31, 38] studied the various aspects of blog reading behavior by
analyzed user log data obtained from Doblog (a Japanese weblog hosting service),
