Tag Archives | Topsy

The Tech That Got People Talking on Twitter in 2010

Topsy is a search engine whose results are determined the pages people are talking about and linking to on Twitter. The more a page gets mentioned–especially by influential Twitter users–the higher it’s likely to show up in Topsy’s results. It’s an interesting approach that works well for stuff that people are chattering about at the moment (strangely enough, the results for “chrome os” are meatier than the ones for “rutherford b. hayes.”) It can also do things like show you a list of Twitter users who are oft-retweeted on a given topic (such as HDTV or cars).

The site is built around a giant database of Tweets that have been analyzed and indexed, so the Topsy folks know a lot about the terms that show up most frequently on Twitter. I asked them to compile a report for me on twenty tech terms and how often they were referenced. It’s not a perfect mirror of the Twitter zeitgeist–Topsy only pays attention to Tweets that contain links and ones which have been retweeted, and it finds keywords even if they’re in a URL rather than the meat of the Tweet–but it’s still a useful reality check.

Here it is…

Continue Reading →

2 comments

A Search Engine Based on Retweets

(Topsy LogoOver at TechCrunch, Michael Arrington has a good review of Topsy, a new search engine that bases its relevance rankings on retweets–the action, on Twitter, of repeating a tweet from another Twitterer to share it with your followers. The idea makes sense–in theory, at least, an item on the Web that lots of folks retweet should be more interesting than one that nobody retweets (or nobody tweeted in the first place). And Topsy gives more value to retweets that come from particularly influential Twitterers. Which is logical for a number of reasons, not the least among them that it helps prevent people from gaming the system.

In its current form, Topsy still feels more like a good idea than one that’s been absolutely nailed–when you perform a search, it comes back with a ton of stuff, and it’s not always sure what’s what. (For instance, in this egosearch for my @harrymccracken twittername, I’m not entirely clear what the organizing principle is behind the list of Twitterers on the right-hand side of the page.) But the potential is huge. Like much of Twitter, retweets are a useful but crude crutch invented by Twitter users, and I suspect that Twitter will replace retweets as we know them with something more elegant. (FriendFeed already has far more sophisticated mechanisms for sharing information and expressing your approval of it.) If Google and other general-purpose search engines aren’t figuring out how to incorporate retweets and other retwe

One comment