NewsGator introduces feed suggestions

NewsGator converted its popular RSS feed aggregation clients to freeware in January 2008, and now that seed has borne fruit: recommended fruit, to be precise.

NewsGator's new Recommended Stories filter introduces users to stories and feeds they aren't already subscribed to, but might like.

(Credit: NewsGator)

Partnering with SenseArray, a collaborative filter from Uprizer Labs, NewsGator now offers live RSS feed recommendations from feeds that the user hasn't already subscribed to. Currently available only on the online NewsGator client, the filter pulls information from NewsGator as well as its sibling desktop clients, FeedDemon for Windows and NetNewsWire for the Mac so that users who synchronize their RSS data will be contributing to the list of recommendations.

Brian Kellner, NewsGator's vice president of products, compared the process to more common Web-based ratings systems. "It takes attention from the client or online site, just like your rating on Amazon, but you're rating it with attention." That attention, he said, comes from marking a post as read--essentially telling the filter that you like it. If enough people do that and the post matches your interest, it might be suggested to you as news you'd like but haven't seen yet.

Kellner said that NewsGator will be making two kinds of recommendations. The first, a general news category, is "wide-open," as he puts it, but limited to posts from the past two days. The second is narrowed down to categories, such as entertainment or sports, and more heavily utilizes the SenseArray filters. These more specific recommendations are limited to not more than a week old, and should be adjustable to the tastes of the user.

The challenge, Kellner added, was how to recommend current stories that users aren't already getting. "We pull in six million events per week that we think are relevant." But the system won't be perfect initially. "Over time, we'll see what adjustments we need to make."

NewsGator does have plans to push the recommendations feature out to its desktop clients, but there isn't a timeline for that, yet.