Refreshed Datasets!

By popular demand, we have refreshed our massive corpus of Twitter data. As part of the facelift, some of our API fields have been eliminated, and many more have been added. Trstrank, for instance, will include a new field called Trstquotient, or TQ, which can be used as a spam indicator. (For details on how that works, stay tuned for a forthcoming blog post). The fields we chose to eliminate from Trstrank–followers, following, and statuses–can be readily accessed via Twitter’s API.

Our new datasets will provide the most accurate and up-to-date reflection of a Twitter user’s measure of influence (Trstrank), activity level (Influencer Metrics), and interactions between two given users (Conversations). The datasets that changed the most, Influencer Metrics and Conversations, have lots of new fields.  Influencer Metrics is now a more rigorous way to measure retweets and @ replies, both incoming and outgoing, and Conversations gives a full summary of the interactions between two users.

Screen shot 2010 08 02 at 8.05.04 PM2 250x300 Refreshed Datasets!

We’re versioning the new API calls, to prevent the unpleasantness that could accompany a rapid switcheroo, but our old calls will be phased out quickly. We welcome your feedback on this exciting update!