Multimedia Social Networks

Background

A multimedia social network is a network in which a group of users share and exchange multimedia content, as well as other resources. Heidemann et al. (2012) provide an overview of online social networks in order to contribute to a better understanding of this worldwide phenomenon. Multimedia social networks can be classified into three categories (Tian et al., 2010):

• •

  Imagery Social Networks: These networks capture the social relationships and activities between users through photos, surveillance videos, or wireless sensors. In such networks, face detection and recognition algorithms are used to recover friend networks from data such as photo albums on the Web. In imagery social networks, a major problem is the privacy problem as it is possible to infer the user’s location or other private information from these data. Litt (2013) explores how background factors, motivations, and social network site experiences relate to people’s use of social network site technology can be exploited to protect their privacy.

• •

  Gaming-Driven Social Networks: These networks allow users maximize their own payoff by exchanging and sharing their resources. Network members watch and learn how others play the game and adjust their own strategies accordingly to achieve effective cooperation. Representative examples include peer-to-peer (P2P) social networks and colluder social networks.

• •

  Interaction-Driven Social Networks: Such networks characterize relationships based on users’ interaction and other activities in online communities. By using interaction-driven social networks, we can effectively develop online social multimedia services such as online video advertising.

By using social multimedia, we can analyze community activity around multimedia resources. We can also derive metadata from social activity and resources, and aggregate data to better reason about their content. In the light of this evidence, De Choudhury et al. (2009) analyze comments on YouTube videos to derive “interests” and topics. Mertens et al. (2006) use community activity for “social navigation” of Web lectures. Shamma et al. (2007) use chat activity in Instant Messenger to reason about the content of shared online videos. Shamma et al. (2010) use the content, volume and trends of Twitter messages about multimedia broadcast to reason about the content of the event. In general, researchers focus on various research activities:

• •

  Visualizing Flickr tags over time (Dubinko et al., 2006).

• •

  Reasoning about Flickr groups (Negoescu et al., 2009)

• •

  Extracting semantics of multimedia tags (Rattenbury et al., 2007) and the relationships between them (Wu et al., 2008).

For social media data, vital key information is required at the semantic level. This information is the so-called 5W’s and 1H, i.e., who, where, when, what, why and how.