I’ve always liked tag clouds. I find that type of data representation useful, not for navigation (although it has its merits), but for quickly understanding the metadata or underlying theme of a given data set. There’s something intrinsically pleasing about being able to grasp data complexity through simplistic visual design. Minimum message length, lowest energy state, Occam’s razor, “keep it simple, stupid” — these terms all come to mind when I think of tag clouds.
I use social media — principally Twitter and to a lesser extent LinkedIn — to engage others with similar interests and to develop and promote my personal brand. As such, I’m always looking for ways to measure my efforts. Specifically, I’ve been searching for a way to generate a tag cloud based on my tweets. A not-so-easy task anymore now that Twitter no longer supports basic authentication over RSS in favor of OAuth.
Earlier this year, Sally Church blogged about word clouds and a new social media service called mirror.me. Mirror.me generates a “reflection” of a person by creating a tag clould of a their tweets. This sounded like a perfect solution for me to get a tag cloud based on my tweets. My top terms include science, health, medical, computational, research, biology, biologist, cancer, bioinformatics and healthcare.
My Tweet stats show that I’ve tweeted 2,399 times since April 25, 2008; 1.5% of my tweets are retweets and 46.2% are @ replies. I average 2.3 tweets a day (last 30 days) and 2 tweets a day (lifetime).
Tweet Cloud is another tool I’ve found that generates a tweet cloud based on a window of time and top number of words, both of which are adjustable. Here’s my Tweet Cloud of the top 90 words covering 2,215 tweets [2,059 (92%) original tweets and 156 (7%) retweets] over the last two-and-a-half years [August 1, 2008 to July 24, 2011]. Interestingly, the stats are different than Mirror.me. My top terms include science, data, research and health.
Another tool I discovered a few months back — albeit not a tag cloud generator but informative nonetheless — is Twitter Statistics, a project on Xefer that uses Yahoo Pipes and the Google Chart API to produce a scatter plot showing a Twitter user’s tweet statistics. Unlike Mirror.me and Tweet Cloud, Twitter Statistics visualizes tweets by hour and day of the week. Evidently, I tweet the most between 10am and 12pm on Tuesdays and Thursdays.
How “on message” are you with social media? How do you measure your personal branding efforts?
Walter Jessen is a digital strategist, writer, web developer and data scientist. You can typically find him behind the screen something with an internet connection.