The what is there, why it is important follows here. When we say xxxx, what do we mean?  Consider the sentence:

Regarding the Swedish discussions about Public Service, not a single one of the ideas suggested by Richard Gatarski sucks.

My point here is not the content of the sentence, but its sentiment. That is, "the attitude, thought, or judgment prompted by feeling". I hope that you find my attitude towards my own ideas are pretty positive. In other words – I do not suck.

In a comment to my previous post Adam Nussdorfer suggests that I should do a Google Suck Index (GSI) with "Gatarski sucks". Of course I did, and currently the result is zero hits. The same goes for the Swedish translation "Gatarski suger". Good enough, eh? The only problem is that soon there will be at least one hit (this post). And if one of you blogs about this post and include "Gatarski sucks", Google will have two hits. Long-tailing around for some time and the GSI for my name would at first sight look terrible. So, I need to rewire your brain behind the current blink at GSI.

In a post in his company’s blog Adam refers to the Business Week article Love the customers who hate you. There Jeff Jarvis, famous not only for his Dell Hell experiences, tells the readers to check their GSI. Read the article, because Jarvis’ advice on what to do with angry customers are insightful and valuable. The same goes for the lead story Consumer vigilantes and the podcast episode behind it.

But Google finds words, not sentiment. Hence you must think beyound hit counts – that is the important lesson here. Regarding my little person it does not matter much, because you (if you are human) can easily scan the few hits to grasp the sentiment(s).

Let us take a more complex, yet illustrative, example. A Google search for (without quotation marks): cheap monday are not comfortable, gives me 2.580.000 hits, and cheap monday are comfortable gives 1.660.000 hits.  Do not assume that means that the majority of the Web stuff about cheap monday’s jeans is about how unconfy they are. Actually, the first 10 hits on the "not" search goes to retailers and fashion promoters. Clearly, we (humans) can not analyze the dominating attitude(s) here by looking at hit counts. We need help.

Machines capable of interpreting feelings and meanings have become more important than they were when I defended my dissertation Artificial Market Actors in 2001. The issue of meanings was not in my focus, although I sort of grasped the importance. In other words I am not the expert here, others are (let us know :-).

As I noted a few posts ago (in Swedish), what most likely will come in handy is "Sentiment analysis". According to (the current) Wikipedia definition it is an:

"area of natural language processing, computational linguistics and text mining. Generally speaking, it aims to determine the attitude of a speaker or a writer with respect to some topic. The attitude may be their judgement or evaluation (see appraisal theory), their affectual state (that is to say, the emotional state of the author when writing) or the intended emotional communication (that is to say, the emotional effect the author wishes to have on the reader)."

For sure not a new field, but something I believe anyone interested in market conversations must learn about and contribute to. Note, this goes far beyond Google and similar search engines. Sentiment analysis, and related fields, are useful for intelligence in all kinds of social and other media.