Brand Niemann, who this year with Mills Davis restarted the Semantic Interoperability Community of Practice (SICoP), at SemTech in June explained the building of the SemTech 2011 agenda and the SemanticWeb.com archive in the cloud with Linked Open Data. The process involved structuring content in MindTouch to create site maps; screen scraping the agenda and articles into Excel; importing data into TIBCO Spotfire self-service BI tool and creating an interactive dashboard from which users can sort/filter/search the data, download it, and share it via Web Player and an iPad app.
“Essentially everything is in the same format and linked underneath so that it is semantically interoperable,” said Niemann, formerly senior enterprise architect at the EPA and now director and senior data scientist at Semanticommunity.net, in a presentation at SemTech.
He describes this in more detail in the video below:
What was accomplished with the SemTech and Semantic Web blog data has a counterpart with another effort Semanticommunity.net undertook with PearlDiver Technologies, Building the Health Data Infrastructure in the Cloud for the Health Data Initiative Forum. This included BuildHealthyPeople.gov in the cloud to make health data more accessible to people at local, state and other levels, and the Medicare Zombie Hunter next-gen medical data mining technology.
Zombie Hunter which is described on the site as “an application that performs live searches through approximately 3/4 of a billion Medicare records to find those pesky dead patients to keep coming back for more medical treatment. Based upon the user input, the Zombie Hunter will identify (to the extent allowed by law and CMS) the undead patients, the patient’s medical provider, the amount of money charged to Medicare, and how much money Medicare reimbursed for the treatment. The data is then visualized by Semantic Community in Spotfire.” As Niemann discussed at SemTech, this led to finding cases to the tunes of many millions of dollars that potentially could indicate waste, fraud or abuse. It’s estimated that some $30 million has spent treating dead Medicare patients. The app provides a visual representation of data in an interactive analytical environment with the ability to filter to isolate specific subsets of data; marking to drill-down to specific data; and roll-up and aggregate data at different levels, without requiring massive hardware to run it.
While they did not win best of exhibit, Niemann says, “ours was the only ‘big data in the cloud’ with semantics apps there. Most were iPhone apps for personal health and wellness.”