Community Search and Socializing the Web

When you search with Google, the results are ranked based on PageRank. When you search within a community search, the results are ranked based on a social metric. For example, let’s say you are in the Ford community and you search on “Sync in Ford Tarsus”, the results set can be ranked against several social metrics like Influencer, where the top results are from the most influential people in your community, or Engagement, where the top results are content with the highest conversation index. But that isn’t the most intriguing aspect of community search. When I think about socializing the web I think of one thing, “giving power back to the people”! What’s popular, what’ not popular, what’s hot, what’s not hot, whose influential or not isn’t determined by some spider crawling the web, or a group of elitist in the web sphere, it’s determined by you, me and the community. Our social interactions help to shape the social context of our data, and community search is a reflection of that. If you search on “Sync in Ford Tarsus” today, and then performed the exact same search tomorrow you will get a different result set because the social interactions and analytics during the past day have reshaped the social context of our data, and the community search results is a reflection of that in the context of what you are currently asking.

Furthermore, the applications like community search, recommendations, connective tagging … etc are not only a representation of our data’s social context, but they are an enabler and influencer as well. For example, community search’s fitness functions will use evolutionary algorithms so that mutations can introduce outliers into the result set, and it’s our social interactions with that data that will influence how that data is perceived, consumed, utilized … etc. This further feeds and matures the social context of our data so that we can say to our clients that we have data that tells the social characteristics and behaviors of their clients, users, learners, employees, students … etc, and we provide the tools to make sense of it.

Social context of our data means our data tells a story about the social interactions, collaborations, influences, engagements, exposure … etc of the community. Instead of scalar values, the data becomes relationship centric, concentrating on the connective aspects it shares with other content/users. This lends itself to easier analysis through mathematical models, such as heuristics, Bayesian, degrees of separations, vector analysis … etc.

Look for me to implement this and more in SportsFlash in the months to come, stay tuned!