Some of you have latched onto the connection between what we’re doing and the Semantic Web. And yes, there is a connection.
Of course, the idea of using the Semantic Web for trend analysis is nothing new, but we believe that many of our processes and algorithms add a refinement that previously wasn’t available through existing aggregate trend analysis services.
We have officially started started publishing limited releases of production data subsets thanks to everyone’s hard work during our initial beta testing phase.
Of course, we still have a long, long way to go, and we won’t be taking upgrades until we are confident of the accuracy of the aggregated data and will also be working on optimizing our algorithms.
We’ll continue to develop more automated testing scripts and refine our testing procedures and quality assurance procedures.
We’re currently fine-tuning the RoboCrush graphing process, including experimenting with various color schemes and ways to make the graphs more readable by the end-user. Once we have enough graphs accumulated we may look at away to do graphic overlaying to find statisically significant overlaps.
We’ve also distributed the mock errors throughout the beta testers so that we can confirm the understanding of the incident naming conventions with the beta testers.
Tad & P.J.