Dygest.net is a fascinating new service that looks over hundreds of topical sources online, finds out what the hot topics are and publishes excerpts of coverage that help tell the whole story succinctly on one page. It's like Cliffs Notes for meme-tracking and it appears to work quite well.
At first glance Dygest looks like the work of one more computer scientist who's trained a robot to hunt for hot links. There are a lot of those now. This one uses natural language processing algorithms instead of tracking links between sources, though, and the summaries from multiple sources are enjoyable to read. Demo sites have been set up for technology, environmental, film, Chicago-area and San Francisco area news.
The digests aren't perfect but they are surpassingly coherent. The site isn't particularly eye catching at all, but as a proof of concept it's very interesting. You can either get a broad understanding of a story by reading highlighted paragraphs from various sources, or pick one of those sources to dive deep into its particular coverage. The opportunity to sample different writing styles, angles and key points makes Dygest fun to read casually and for in-depth exploration of the news.
The team behind Dygest says they think the service would be well suited for mobile consumption, but white labeling the technology is high on their list of ideas as well. OneSpot comes to mind as an immediate competitor, but there are others.
Dygest has pre-seeded each of its demo sites with a few hundred sources, but the possibilities the analysis engine presents are many.