Blog

Looking at the words and phrases that people use in social media and then using sentiment analysis to assess how positively or negatively they feel about something. With Facebook, the opportunity is huge. If you combine the ability to analyse the sentiment in status updates with the vast amount of profiling data, the potential for insight into consumer behaviour is huge.

Sure it is.

The thought that anyone could seriously believe that it is rational to use the status updates or wall postings of Facebook members to gauge wholesale happiness (and then slice & dice it demographically), appalls me.

This is like grabbing a dictionary throwing it on the floor and pointing to a word and capturing it, doing this repeatedly and calling the end result literature.

Capturing random, out of context, variously dated snippets of text is not the raw material for any sort of serious research.

Facebook postings can be (and are) written seriously, for marketing, sarcastically, or just be random top of mind musings (I like shoes; My cat is hilarious; I hate mayo; etc, etc). I'm sure there are a host of other reasons people update statuses. NONE, are useful for gauging widespread "happiness" or any other sentiment.

Text analytics and semantic interpretation of web content is being done by many people; some of it is quite good. This is not.

Sound research requires some sound methodology.