SO question on sentiment analysis — from 2009

“Is it possible to guess a user’s mood based on the structure of text?”

Very interesting to read this Stack Overflow question today. The question and responses, written mostly in 2009, really bring to light how much public awareness of this field has grown in recent years. The same question would likely not be asked today — a Google search would inform the curious and thus questions on the topic would be more specific than this one. If it were asked, the answers would be quite different.

I especially enjoyed looking at the train of thought that many followed in trying to answer the question. Some thought the problem was completely intractable, others that text in itself was insufficient (“…but if you had keystroke timings you should be able to figure it out”). Those that suggested solutions focused on counting occurrences of emotional words (“maybe you can count the number of 4 letter words?”) or punctuation in the text with various manually-assigned weights, or ruminated about the use of ALL CAPS or font styles (“bold red text is probably an angry user”). The use of spelling and grammar mistakes was suggested as some type of indicator — were these mistakes really so rare in 2009 that they could be interpreted as emotional?

(I don’t mean in any way to be derogatory towards those that chimed in, of course. )

One of the answers (the one most up-voted) used the term ‘sentiment analysis’ and described a solution schema that is still used effectively today. Which points out for us that the requisite level of research in NLP, as many of you already know, has been around since 2009 and long before 2009… it is public interest and awareness, and the sheer number of people interested in these types of questions, that has grown so dramatically since.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s