For the developer seeking to experiment efficiently with social APIs, O'Reilly's 2nd Edition of "Mining the Social Web" is a truly outstanding resource.
Author Matthew A. Russell drops the developer right into the sandbox of each social network (Twitter, Facebook, LinkedIn and Google+ are particularly emphasized, as you would expect) with just the right amount of explanation about what's accessible via each dataset, and then clears out all the obstacles so they can start data mining against very clear examples.
Included with purchase is access to a virtual machine "that has iPython Notebook and all of the other dependencies that you'll need to follow along with the examples from [the] book preinstalled and ready to go."
Russell is transparent about his preference for the Python programming language as a teaching tool in this social mining context, and one Amazon reviewer (Bernard Enjolras) goes so far as to call the book "the best introduction to real-world programming in Python."
Wherever you land in religious discussions about programming language preferences, you will appreciate the quality of Russell's writing, and how concisely he sets the tone for mining each social network before dropping you into code examples. Chapters make no assumptions of the reader's existing knowledge about the social network in question, but Russell manages to set the stage without the breathlessness that's so common in books about social media.
For example, the chapter on "Mining Twitter" follows this progression: 1.1. Overview, 1.2. Why is Twitter All the Rage?, 1.3. Exploring Twitter's API, 1.3.1 Fundamental Twitter Terminology, 1.3.2. Creating a Twitter API Connection, 1.3.3. Exploring Trending Topics, 1.3.4. Searching for Tweets, 1.4. Analyzing the 140 Characters, 1.4.1 Extracting Tweet Entities, 1.4.2 Analyzing Tweets and Tweet Entities with Frequency Analysis, 1.4.3., Computing the Lexical Diversity of Tweets, 1.4.4. Examining Patterns in Retweets, 1.4.5. Visualizing Frequency Data with Histograms.
When you've gotten your sea legs mining Twitter, Facebook, LinkedIn and Google+, you'll find that the book still has a lot more to offer in more advanced concepts related to natural language processing, web scraping and parsing, and the n- number of technical challenges that arise when trying to apply order to human language data. The book goes beyond the most popular social networks to also focus on "Mining Mailboxes" and "Mining GitHub." The final chapter ("Twitter Cookbook") includes a "Problem / Solution" approach to 20+ other opportunities to streamline your efforts in mining Twitter.
We think it's a worthy inclusion on the bookshelf for any developer learning PHP, learning social media integrations, or venturing into social data mining.