Unlike other traditional news sites, Fark’s appeal is its very ambiguous headlines, such as “High School math teacher arrested on ‘What is 210 = 210?’” While this article is clearly about high school math teachers, it’s unclear that this article is about “marijuana," making it perhaps difficult for Fark to broaden the consumption of that particular article by people who would have been interested in "marijuana" and not necessarily math teachers.
“People who read Fark know the taglines almost always have nothing to do with the article, said Drew Curtis, founder of Fark, during a recent interview with me. “They’re [headlines] at best vague and at worst, deceptive intentionally,”
The challenge for Fark, therefore has been to broaden the consumption of articles. One solution was to categorize its content under topics in which they could be discovered. Enter Digger.
“It would take an army of editors to do this manually,” said Tim Musgrove, founder and CEO of Digger, a semantic search engine, that helps automate the tagging and categorization of content for publishers. “In the case of this article, our tool is supposed to determine what the article is about, and even if it says Mary Jane, our tool is supposed to determine what this article is mainly about.”
The new categorization will be going live on Fark sometime next week. At that time, Farkers will be able to search for topics, and find a list of articles that are about that topic, something Fark readers cannot do today.
It's unclear how much pageview lift Fark will get from the new categorization of its content. Both Tim and Drew wouldn’t disclose what kind of pageview lift they expected, but Tim did say that results found from the new search might get click-through-rates in the teens. It's also unclear whether Fark will be able to monetize these new pages as they will clearly be targeted to a smaller slice of the Fark population.
Fark isn't the only publisher embracing semantic technology to organize content, HuffPost is also using semantic technology to reorganize its content to create local versions of HuffPost. For more on Digger's clients, watch this earlier interview with Tim - Digger helps categorize and tag content.