A new Facebook feature scans users’ posts and then groups messages around particular topic pages
Facebook on Monday added a news feed that relies on natural language parsing to “read” what contacts are commenting about and group their posts around a topic-related Facebook Page.
The feature builds on previous aggregated news feeds that show when contacts linked to the same web page, checked into the same “Place” or wrote on the same contact’s Wall. Those features, however, do not require the natural language processing used in the new feed.
The new feature groups posts that refer to a topic that has an existing Facebook Page related to it, for example a film or product. Users can explicitly link their post to such a Page using the “@” symbol, but the new aggregated feed includes posts that mention the topic in question even if they have not chosen to create a link themselves.
The feed creates a way for Facebook to drive traffic to Pages without needing to modify users’ posts to include a link. It is intended to operate in a way similar to Twitter’s trending-topics feature, which lists popular topics grouped by keyword.
The company confirmed to specialist publication Inside Facebook that the feature uses natural language parsing, but doesn’t attempt to identify the sentiment behind the post, or whether the term involved is the primary topic of the post. That means, for instance, that the posts grouped around any given Page could include both negative and positive comments as well as posts that aren’t necessarily primarily about that topic.
Facebook first tested the feature late last year with a feature aggregating posts that “mentioned” a Page without explicitly tagging it.
The feature is likely to draw criticism from some quarters for “reading” users’ status updates, somewhat as email providers including Yahoo and Google have been criticised on privacy grounds for scanning users’ emails in order to deliver targeted advertising.
In a similar way Facebook has been criticised for its rollout of face-recognition technology that scans photos to suggest which individuals are included in a given image.
The feature, which was rolled out in the US in December 2010 and extended outside the US in June, uses facial recognition software to match new photos to other photos users are tagged in. Similar photos are then grouped together, with Facebook suggesting the name of the friend in photos.
German data-protection officials recently requested that Facebook disable its facial-recognition software and delete any previously stored data. Making facial-recognition technology opt-out runs afoul of European and German data-protection laws, John Caspar, Hamburg, Germany’s commissioner for data protection and freedom of information, said in a letter to Facebook on 2 August.
Meanwhile, at the recent Black Hat security conference a Carnegie Mellon University researcher used Facebook photos to demonstrate how facial-recognition technology can be used to identify people as they walk down the street. Using off-the-shelf facial-recognition software and students’ photos posted on Facebook, Alessandro Acquisiti, a CMU researcher, showed attendees at the annual Black Hat security conference how he was able to positively identify 30 percent of students walking around campus.