The breaking news is that a new service from Regator has resulted from tweaking its semantic algorithms to find within its human-curated collection of web content emerging stories and to quickly alert bloggers and journalists about them via a desktop app.
Regator actually has been around as a curated blog directory and search engine for a couple of years, and in and of itself is a perfectly good and pretty fast source for the word on the digital street, along with the other usual suspects (Twitter, Facebook, Google blogs, CNN, etc.) – at least in so far as the big stories go. “But unless it’s a really big story and Twitter explodes eventually, you won’t find those second-tier news stories so easily,” says Scott Lockhart, Regator cofounder and CEO.
Those more specialized or less broadly covered stories can be bread-and-butter for a lot of online writers, and knowing about them sooner – without having to hunt through RSS feeds, search hashtags, or troll through other news sites – can present reporters with some of their best opportunities for their articles to be among the first wave of coverage. The same can go for some of the bigger stories, too – for instance, Regulator Breaking News had the French embassy convoy Baghdad bombing story up before it was covered by CNN today.
The Regator desktop app that provides these alerts lets users click through to a page that compiles recent posts on the news, images, Twitter activity and historical coverage that relates to the story, such as the subjects mentioned in it. Its trends capability, as Lockhart explains, goes beyond the more common approach of deriving results off of keyword-density searches, which can generate questions over reliability and consistency.
“What we do is use a complex algorithm that is a combination of things [math-driven as well as some NLP processes] to really look at what is happening through thousands of blog posts a day, identify those key terms that are important, and look at the relationship of those terms to other terms,” he says. Say, as a for instance, Anthony Weiner’s association to John Boehner’s, which recently was trending in the context of the latter’s call for the congressman to resign. Lockhart says Regator did some playing around with the algorithms that let it do medium- or konger-term trending so that it can raise a quick alert that something is happening now to enable the RBN service. “It really does alert you a lot faster than things coming up on Twitter,” he says.
Lockhart says one of the advantages of what Regator can do with its breaking news service goes back to the value of the data it’s culling from, drawing as it does from a hand-selected batch of blogs. Only about 18 percent of the blogs that are nominated for inclusion on its site are approved. With a lot of other similar services, he says, the algorithm is focused on cleaning up the data, “but we start with really good data. …When you combine human curation and semantic algorithms you’re really onto something.” He expects that future capabilities will include more classification, so that searches users might want to do on the main site or the breaking news service can respond beyond the specific query – for instance, pulling up an iOS 5-oriented trend or blog for a search on iPhones.
The RBN service is the second of Regator’s B-to-B plays. It also has a commercial semantic API that it offers to businesses to help them improve their internal data with categorization and more metadata for aggregation and search purposes.
- Airomo Launches Appcurl for the Amazon Kindle Fire and Barnes & Noble Nook Tablet
- Jinni Launches "Watch Together" Social Recommendations
- Digirati and Semantic Web Company Innovate Semantic Publishing
- Big Data & Semantic Search: How They're Affecting the Job Market