In the following example, latest trends in loud speaker development are discovered and tracked. Initially, a degree-of-separation Google query for "loudspeakers audio forum" is run, returning the following top ranked sites (Figure 13).

Figure 13. Results to "loudspeakers audio forum" degree-of-separation search.

One of the top-ranked Web sites is ecoustics.com. The "ecoustic.com" forum is now chosen for an in-depth analysis. The forum consist of different threads, shown in figure 14.

Figure 14. Overview of different threads in ecoustics forum about "Speakers"

Figure 15 displays the different attributes of interaction that can be extracted from each thread, such as the thread name, the initiator of the thread, the nickname of each poster as well as her or his username, the timestamp of the post, and the contents. A social relationship between two people is constructed if one poster responds to the previous post.

Figure 15. Information collected from messages posted in thread

Parsing and collecting all threads about "speakers" permits to create a social network of all participants to find influencers and gatekeepers as well as analyze the contents of their discussion.

Figure 16. Social Structure of 4 months of discussion of speaker thread of ecoustics forum (x-axis=time, y-axis=actors, z-axis=actor betweenness)

Figure 16. shows the social structure of four months worth of interaction in the thread about loud speakers in the ecoustics forum. It is easy to see that there is a small group of very central people dominating the discussion, and the bulk of the participants joining and leaving the discussion in the third month.