– Relating to meaning in language or logic
One of the new(ish) buzzwords around town seems to be the ‘semantic web’ or web 3.0 as some are calling it. I’ll be honest and say I’m only ankle deep in the whole thing, but as far as I’m aware, the idea is for a web with meaning attached to the data. So, rather than search engines simply returning the results of keywords, those words will have meaning and relationships. Using metadata, sites and software will know the relationships between names and words, and manipulate the content in a useful way.
That’s the theory anyway. One of the services aiming to deliver a semantic web is Twine by Radar Networks. I read an article about Twine on Think Artificial a few weeks ago and tonight I thought I’d take it for a spin. Twine describes itself as a “social interests tool” rather than a social networking tool. It’s a way to keep up to date with your interests, rather than your friends (although it has some social networking capabilities too apparently). Twine allows you to gather all the data from your travels around the web (bookmarks, images, videos etc) in one place and filter them semantically. It’s early days so some teething issues are to be expected, but overall I have to say my initial impression of Twine is less than favourable.
I’ve never really gone in for bookmarking services like Deli.cio.us or social networking services like Facebook, so I don’t really know what I was expecting. But since my bookmarks are in need of a good spring clean, the allure of a system that learns your tastes and allows meaningful searching of data was too much to resist.
In practice I found the interface and it’s manipulation very basic and clunky. You create twines on topics that interest you and Twine constantly returns the latest submissions to that area of interest. Twine seemed to do that okay, but I was disappointed to find I couldn’t customise the twines homepage much (I was looking for an iGoogle like drag and drop functionality) so having multiple twines seems like it would be a bit cumbersome. Also, adding items to Twine could have been more intuitive. Sure, if you find a website you like you can just click the Add to Twine bookmark that they give you and a dialog pops up, allowing you to name the bookmark etc. The problem was, I accidentally added the item to My Items and an existing twine on a completely unrelated subject (some sharing option that it wouldn’t let me deselect). Since it seems you can’t remove items from twines you don’t manage, I couldn’t see a way to correct my mistake.
Another source of confusion was the functionality for importing your bookmarks. While I knew that sorting out my bookmarks in Twine would still mean going back to my old bookmarks folder and manually having a clear out, I thought that Twine would be a good way of filtering them easily to begin with. In practice though the tagging filters were next to useless: simply filtering my bookmarks into two categories – “bookmarks” and “videos” – and coming up with nonsensical related tags. I suspect this is because, as some reviewers have said, Twine relies on properly structured webpages to allow it to extract the metadata, and unfortunately the web is a scrap heap of poorly made webpages, so I can’t blame Twine completely there. However, even with my bookmarks in Twine, I didn’t seem to have basic options like batch commands (multiple item delete etc) or alphabetical sorting, and the lack of effective filters made clicking through individual bookmarks a chore. In the end I gave up.
That said, partly because I can’t figure out a way to quit Twine, but also because I should give it the benefit of the doubt, I’m going to persevere a little longer. I still like the idea of semantic tagging of items and a more intelligent web, I just haven’t found it so helpful in practice. I have a suspicion that Twine may not be for me after all, but I’d be interested to hear other people’s thoughts on Twine or similar services. In my case, I’m basically looking for a way to keep up with my interests with an rss-like delivery. Other features like recommendations and social networking would be a bonus. Twine promises to deliver much of this but so far, it seems, not very intuitively.