As the ways to communicate and share information online multiply, so do the headaches of keeping track of all that information. Let me see, did I store that information as a browser bookmark? In Facebook? Did I tag it? Was it on someone’s blog? Or was it in an e-mail? What about pictures and videos? How can I store and find them when I need them?
A product called Twine, currently in beta testing, aims to help users lasso those disparate threads and make them available for easy retrieval and sharing. Using a combination of the “semantic web” (for more on the semantic web, see Duane Degler’s What You Should Know About the Semantic Web, User Experience magazine, Volume 6, Issue 2, 2007), natural language processing, and machine learning, it automatically tags and stores information for the user in multiple ways, and over time learns the user’s habits and preferences to constantly improve its effectiveness.
After putting a special toolbar on your desktop, you can use what the product calls a bookmarklet to create and add to twines on any subject. The twines then gather information from all the different sources you have looked at to create a one-stop shop for information retrieval. You can also make your twines available to friends, and they can add to the information in a shared twine. One challenge, still being worked on by the company, is to avoid duplication of content when twines are shared.
What makes this offering particularly interesting from the viewpoint of time-stressed, info-overloaded users, is that the program auto-tags the material you store, summarizing and categorizing it. At any time, you (and your friends) can re-sort the information any way you want: by the most recently changed information, by people most closely connected to you, by interests, or by whichever arrangement makes sense to you.
Nova Spivak, founder of Radar Networks, the company developing Twine, describes it as “moving from a file server metaphor to a database metaphor.”
Company user experience specialist Kim Laama notes that one major usability hurdle has been merely to explain the concept in a way that people will understand. When the company ran usability tests with fifteen users late last year, they found that even in tech-savvy San Francisco, jargon can be off-putting.
“We got a lot of specific feedback on pages and features, but feedback on our messaging and the bookmarklet was the most significant. We improved our messaging to be less reliant on buzzwords, like Web 3.0 and semantic web,” said Laama. “We needed to find a way to communicate in a less technical way but still answer the questions, ‘What is twine?’ and ’Why is it valuable?’” She continues, “We are still refining our messaging today. It’s a challenge to demonstrate the value of the semantic web without actually talking about its capabilities and potential.”
User testing has also surfaced challenges of striking the right balance among the needs of different types of users. While some want the many options originally provided by the bookmarklet, others want a one-click experience. The company is now trying to find ways to meet the needs of both types of user.
Throughout the product’s development, Radar Networks has relied on a fast-paced iterative design process. It is now supplementing that process with massive beta testing. Clearly there is interest in what they are trying to do.; according to Spivak, there’s a waiting list of some 30,000 people eager to join the beta testing.