by Selene Arrazolo
The Data Day Texas conference in Austin once again brought together hundreds of folks in the Texas big data community for one data-filled day. Here are the presentations that we thought were “best of show.”
Lisa Green, Director of Common Crawl, discussed the challenges of capturing open data at web scale. Her talk touched on the difficulty behind the lack of standards for formatting and interoperability as well as the amazing wealth of open data now available online.
Andrew Trask and David Gilmore from Digital Reasoning gave a compelling presentation on Deep Learning. They talked about the system they’re building to tackle common natural language processing (NLP) obstacles around syntactic, lexical, semantic, and contextual issues. These can lead to breakthroughs in areas like sentiment analysis.
Keith Casey from Clarify discussed what sets them apart and how they make audio analysis scalable. Audio search is one of the many niche search opportunities yet to be conquered. It’s also one of the most intriguing given the volume and value of data stored in audio files. Identifying and retrieving content embedded in apps, images, and video files is likely to be an area of intense activity over the next few years.
Idibon‘s CEO, Robert Munro, presented Building Better Experts: the co-optimization of human and machine intelligence. He examined why the “cost of human processing has remained unchanged and remains an expensive task,” emphasizing human engagement as an indispensable resource. Idibon’s staff includes subject matter experts, linguists, and crowdsourcing managers.
posted by Shyamali Ghosh on January 26, 2015
by Matt Manning
The recent news that Dun & Bradstreet purchased data-dealer turned data management service provider NetProspex further advances two seemingly contradictory corporate storylines for the B2B data industry’s bellwether. This may be another in D&B’s string of acquisitions designed to buy the company’s way into the future. Or, this could be a case of D&B eliminating competitors who resell competitive sources of baseline company data.
So which is it?
Previous D&B acquisitions like Hoover’s, Fliptop, and Indicee were clearly designed to buy revenues to add to the company’s top line while strengthening what the company calls “modern” sales channels (CD-ROM, the Internet, social, and the cloud). Revenues from these acquisitions weren’t huge, so they would appear to be more strategic in nature than born of any great concern that they would eat into the firm’s base of subscription and licensing revenues.
Over the last couple of years, D&B has also undertaken broad strategic efforts to maintain and extend the company’s relevance:
- incentivized vertically oriented B2B information services to append DUNS numbers via the D&B Exchange;
- made it easier to connect D&B datafeeds to marketing automation platforms like Eloqua;
- added the ability to purchase D&B data to CRM platforms like Salesforce.com (via a deal with Data.com) and SugarCRM;
- reached out to strategic analytics partners like FirstRain and LatticeEngines; and,
- incorporated reliable, international, open source data via a licensing deal with OpenCorporates.com
So will D&B’s many efforts to turn their ocean liner of data succeed? On the face of it, they’re doing all the right things. They’re moving into new delivery channels (almost always via an acquisition). They’re striking distribution deals with most of the biggest marketing data movers and shakers. They’re adding more international data. But what about their data itself? Will their formidable resources now focus on strengthening their core proprietary database? Or will they sidestep (once again) the elephant in the room and continue to build out a robust distribution ecosystem without attending to the cracks in their database business’s foundation that undermine all these investments?
posted by Shyamali Ghosh on January 12, 2015
