Detecting sentiment in content on the web, including Tweets and news articles is becoming an increasingly popular way of interpreting data. With that it mind, RankSpeed is launching a search tool that does a sentiment analysis of Tweets and blogs.
RankSpeed lets you search for any keyword or tag on a website and attach a sentiment to the search. So you can use the emotional concepts of good, useful, easy, secure, and RankSpeed will then give you search results of blog posts that match the subject and sentiment. For each result, the engine computes the percentage of bloggers that have said a particular product or site is good or easy or useful. If it sounds confusing, that’s because it is. It’s a little easier to understand when you take a look at the site.
You can also filter your search by blogs that are considered (by RankSpeed) to be more useful, easy and so on. A website is ranked as more useful than another if its name is more often associated with “useful”, “helpful” or other synonyms in blog posts and tweets.
For example, I did a search for “good” Twitter Apps. The top result was Identi.ca, a microblogging site similar to Twitter. Hmmm, that’s not quite a perfect match for “good” Twitter Apps. But when you do a search for “popular” Twitter Apps, the top results include Tweetdeck, UberTwitter and other common Twitter apps.
While RankSpeed says its search engine tracks 3 million websites and blogs, it’s obvious that the site’s technology is not perfect by any means. Another feature that is missing is the ability to see the “sources” (or blogs and Tweets) of the results. If I saw that Tweetdeck was rated as the most popular Twitter app in the search results, I probably would want to see who found it popular in order to validate the results. Although the site provides an interesting way to search blogs and Tweets for sentiment-driven keywords, it won’t replace Google anytime soon.