SAS wants to crawl social chatter

Analytics software company releases social media analytics tool for enterprises keen on keeping an eye on what's being said about them online.
Written by Victoria Ho, Contributor

SAS has released a social media analytics tool for enterprises keen on observing online chatter about their business through a magnifying glass.

The analytics software provider announced on Monday a new product called SAS Social Media Analytics, targeted at medium and large companies. The software will archive and analyze social media "conversations" from platforms such as Twitter, YouTube, forums and blogs, according to SAS.

ZDNet Asia understands this information will be culled from public streams of data, not private feeds. For instance, information crawled from Facebook will be from the enterprise's fan page, not individuals' walls, explained SAS customer intelligence product manager, John Bastone, in an e-mail interview with ZDNet Asia.

Bastone said the data would be presented in two types of interfaces--one for office workers interested in analyzing information at a glance, and another more indepth dashboard for marketing analysts within the company.

This is aimed at sifting the flood of data that will be analyzed, he said, noting that the software will perform text analytics with linguistic-based rules in order to predict and interpret semantic data. He used an example relating to a TV retailer, where words such as "resolution", "black levels" and "contrast" all pertain to an overall concept of picture quality.

"You can reprocess that data to account for those words so that you have today's definition of picture quality, applied to historical [descriptive words]," he said.

Upon implementation, the software will comb information dating back two years from the date of installation. The product is delivered on an on-demand model but Bastone said this is not a "pure software-as-a-service (SaaS) offering", noting that enterprise customers could choose to bring data into the company's walls.

Editorial standards