Classifying lifestyle content
Scripps Networks has chosen a text-mining solution from Nstein Technologies to semantically analyze its immense library of lifestyle content. Scripps Networks owns and creates content for five home, food and lifestyle cable networks and Internet sites, including HGTV, Food Network, DIY Network, Fine Living Network (FLN) and Great American Country (GAC).
With its sites recording an average of more than 18.5 visitors per month, the SN (Scripps Networks) Digital team wanted to enhance the functionally of its search technology, and to create a taxonomy and thesauri based on the top terms searched. It chose Nstein to incorporate those foundational elements in its semantic analysis engine.
"All good search is based on the quality of the tags," says Michael Campbell, program manager of search for SN Digital. "Without proper tagging and a broad and deep tagging exercise, it really can’t be an optimal user experience."
As an example of the frustration users previously encountered, Campbell says, "The HGTV Dream Home is exceedingly popular, and is often a search term used by site visitors. But instead of always getting the most recent information and content, they could get something from years prior. Another example relates to our food sites, where terms like ‘dressing’ could related to salad dressing or stuffing and needs to provide both options. We want a search that gives users what they came to the site to find."
According to a recent press release from Nstein, its TME (Text Mining Engine) allows a combination of machine learning and business rules to automate much of the tagging with the highest degree of accuracy possible. Editors still have the ability to verify results, but the overall burden of manual tagging is greatly reduced, Nstein says.