The natural arc many of my clients have followed when it comes to “social media” involve getting:
- a presence on Facebook, Twitter and other social websites,
- friends and followers of the brand
- social interactions on their own websites, such as Like button clicks, and finally
- actionable data on everyone to sell them new stuff.
Some clients are further along the arc than others but everyone is looking to that final segment to bring in revenue and make it all worthwhile. (They aren’t making friends on Facebook just because it’s fun!)
Harnessing social data is a tricky proposition because the APIs and methods for capturing this data are still in the formative stages, even though they are maturing quite well. Facebook’s Graph API and embrace of Open Graph offers a lot but privacy, authentication and the changing nature of their API makes developing applications that handle Facebook data a little tricky. Twitter data is easier to access but they have had less time to develop their API and don’t have to worry about privacy much since they are essentially a publishing service.
Matthew Russell‘s Mining the Social Web knows its stuff and covers several social media websites and the data they provide. Some sites are major players, like Facebook, Twitter and LinkedIn; others that are covered, such as the Google Buzz service, are not as well-known and don’t usually come up in social media discussions. It’s amazing when you think about Buzz not being well-known, because every Gmail account includes a Buzz account. And if you’re surprised to see Buzz in a social media book, then consider the chapters on blogs, email and even microformats, which are important to discussions on data but are not considered your typical social media service.
Mining the Social Web is not like other books I’ve reviewed: it’s more for data analysts than developers, though there is a lot of code and development going on in its pages. Matthew’s language of choice is Python, which is intuitive and an easy for all readers to explore but isn’t common for web developers. But I got out of my comfort zone and learned several things from this book, so I recommend it to all readers who need to improve their understanding of data.
One more difficulty 21 Recipes for Mining Twitter and Data Source Handbook. I’m looking forward to reviewing these books to see how they will illustrate the full scope of social media data mining, not just for data but also for development.