In the latest round of funding from the Knight Foundation’s Prototype fund, two projects focusing on metadata received US$35,000 (£22,000) and professional mentorship to help reach their objectives.
One is Contextly, a company that acts as an engagement service for publishers. Co-founder Ryan Singel, who previously wrote for Wired, said the project aims to “get people deeper into a publication” by recommending more personalised stories.
Contextly will analyse a story’s metadata, like the author or publishing date, and find keywords from the article body that enable computers to understand the meaning and structure of the published information.
Each article’s HTML code contains individual tags that make it unique, but creating these tags can sometimes be a complicated process for publishers.
Over the next four months, Contextly plans to simplify the process of adding metadata to stories by developing what Singel calls a “near real-time writing interface”.
This would provide useful information to writers as they are doing their work by identifying when they have added a significantly new chunk of text.
“Showing new possible metadata candidates in a panel on the side, while a writer is trying to craft a perfect sentence, could just be distracting,” Singel said, so one of the challenges they face is “developing a way of making metadata suggestions a continuous process that runs in the background”.
The platform will have an open-source structure, allowing publishers to build on it by bringing forward their own contribution.
More like this