Thursday 12 March 2015
I was intrigued (for about five minutes) today when I was sent a link to a new IBM Watson demo http://watson-um-demo.mybluemix.net/
The site allows you to enter some of your prose, fiction or non-fiction, and then analyses it to determine the characteristics of the author and the sentiment of the article.
The example had a mind numbing ramble about a guy called Ishmael which then got ranked and assessed to reveal that Ishmael is “inner-directed and sceptical”, “is independent”, “has a strong desire to have time to yourself” etc… etc…
I came up with my own assessment having had to poke myself in the eye a couple of times to keep myself awake after the first couple of hundred words. It consisted of less words but I’ll keep that to myself…
I marvelled at the charts and insight that was gleaned from only (yawn) 2190 words. Try it out and take a look!
I did (eyes starting to hurt… poke poke).
I put some of my past blogs in it.
Apparently when I wrote some of them I came out as… “inner-directed and sceptical”, “is independent”, “has a strong desire to have time to yourself”.
This is all a bit like reading your stars in the daily rag or having a tea leaf reading at Butlins.
Nice and vague but has enough in there to pique your interest – a few niceties that make you feel good about yourself, a few “hmmmm perhaps I might be like that” and a few “… what a load of pants”.
So what’s my point?
I think this is really interesting technology. The ability to make sense out of the written word through software is certainly worth the effort.
Analysing sentiment, analysing motive, positivity and negativity are all the more important in the world of big data where there is far too much data spilling through the ether to be checked manually…
I wrote about this a while ago… https://www.connexica.com/panning_for_gold/#sthash.i6HWCONQ.dpbs
Hold on a minute, what does Watson think about that blog?
You are heartfelt, somewhat insensitive and skeptical.
You are empathetic: you feel what others feel and are compassionate towards them. You are self-assured: you tend to feel calm and self-assured. And you are philosophical: you are open to and intrigued by new ideas and love to explore them.
Your choices are driven by a desire for modernity.
You are relatively unconcerned with tradition: you care more about making your own path than following what others have done. You consider independence to guide a large part of what you do: you like to set your own goals to decide how to best achieve them.
So here I am on a plate. I’m insensitive and sceptical! I’m compassionate and intrigued by new ideas…
Let’s hope version two makes more sense!
If you’re a HR Manager don’t use this as a way of filtering your candidates or trying to determine the likelihood of them being axe wielding murders or simply too “sceptical or insensitive” to employ.
Speak with them and make your own impression. Write down your thoughts and then throw them into http://watson-um-demo.mybluemix.net/…
We at Connexica look at sentiment analysis too but we don’t try to make quite so many assumptions about the prose. We have been looking at the Stanford Sentiment Analysis algorithms which look at the positivity or negativity of text. http://nlp.stanford.edu:8080/sentiment/rntnDemo.html
So Sherlock, which is better? For me it’s a work in progress but definitely a step in the right direction!