Let’s look at the sad, angry data.
On Valentine’s Day, my friend Sean Peter Drohan started a Google spreadsheet where he encouraged people to fill in the texts they wanted to send their exes. Under the cover of anonymity, a social community was born. People bared their souls — playfully and poetically.
Next to the column where users entered the text they wished they could send, people added their reactions.

To maintain the sacredness of the spreadsheet, I won’t reveal which text(s) are mine. But I was intrigued by the curious little data set that had been aggregated: hundreds of texts that a group of people want to send their former lovers, but never will. So I analyzed them.
Here, as of 10:30am on February 15, are the results of my analysis.
1. “You” and “I” were the most-used words.

There’s something tragically beautiful and sad about the fact that “you” and “I” appear next to each other in the data set. Interestingly, my Text Analyzer data revealed a 7% lift in appearances of “you” over “I,” suggesting our broken hearts are more likely to refer to our ex than to ourselves in the texts.
2. The language was most likely to be sad, angry or tentative.

Using a linguistic analysis tool built on the IBM Watson Developer Cloud, I analyzed all of the texts to detect the dominant emotions, language styles and social tendencies. While the tool is meant to analyze text from a single author, and emotion and sentiment analysis is still a growing discipline, there’s still some insight here.














Leave a reply