Challenges of Triaging Twitter feeds

Having used my iPad for over a month now, I should say that the Flipboard app is the most interesting one on the device. It turns my Twitter account into a beautiful magazine that I enjoy reading every tweet that is pulled from my twitter client.

Deciding which of the several tweets would be of interest to me has always been on top of my mind. After all, who would be interested in reading junk!

Few ways by which you can Triage your tweets are as follows:

1. In almost all the cases, the topic of the article pointed to by the tweet is evident from the title itself. So, I can decide and filter almost all of my tweets based on the tweet headline. While this is a good strategy, it doesn’t work always because the complete clue might not be evident from the title itself. Imagine someone replying to his friend saying “This article refutes your assumption!”. You definitely can’t say what is it referring to, unless you have tracked the conversation of this tweet.

2. Everyone is aware of Twitter lists, a group of selected twitterers put together in form of a list that can be followed by anyone. No doubt this is an organized way of segregating a particular Twitter topic/trend, it relies upon the basic assumption that the people in that list are consistent with the kind of topic they are posting.

3. HashTags are always there to attach a tweet to a particular tag that may be searched for later. It does help in grouping tweets together so that a simple search for the particular tag will show all those tweets which are tagged with that HashTag. I think for the search to be effective, first of all, all tweets should be tagged and also the tags should form a closed well defined set. In addition, twitterers should be aware of the best HashTag before tweeting. As far as I know, none of these are in place yet.

4. Twitter trends according to me, is too basic and naive. I am not sure how effective is a search based on the Twitter trend. I am not aware of the way it is constructed and the trends themselves are not a closed set. To add to that, the trends are subjective.

While there are few players who help you search for relevant content from Twitter, I am still not aware of the most accurate engine in this space. Anyone into semantics and contexts will easily understand the challenges of deducing the meaning from a small context of length 140 words!

If there is someone who can really make a dent into this space and at the same time cater to the usability part, it is the combination of Flipboard and Ellerdale. Flipboard no doubt is the best app for anyone who enjoys reading tweets in the form of a beautiful magazine, all they need now is to add some semantics into their engine. Ellerdale’s algorithms have been extensively used in different search engines. We all know how good iPad is for consumption of content and if these two guys can crack the challenges of the list above, I think Flipboard with Ellerdale’s capability inside its magazine will be the killer app for iPad. It would be interesting to know if there are any other challenges besides the one listed above and in any case, how does Flipboard leverage Ellerdale’s capabilities in Triaging tweets!

Waiting anxiously for Flipboard to release their semantic power!

Advertisement