Mashable recently wrote a piece about “vertical-specific” social media monitoring, and whether it could be the market’s next big thing.

A company called Cruvee who specialize in monitoring the wine industry were featured, and there have been a few other blog posts about them lately too.

The main reasoning is that a tool built specifically for a particular industry is going to offer much stronger and more accurate linguistic analysis of that industry than catchall, universal tools.

This reminded me of a part of our sentiment analysis that is sometimes not made fully clear.

Not a new concept…

We are very aware of the importance of industry-specific language here at Brandwatch and we do our best to offer language analysis that specializes in industries as much as possible.

We constantly refine our language systems by adding newly trained classifiers (a classifier is the particular system used to detect and analyze the language of a query’s matches – which classifier should be used is determined upon query creation).

We have over 500 classifiers for different industries across the 17 languages we cover.

Educating sentiment machines

The larger and more specific the corpus is that the classifiers are trained on (we use machine learning to train them – see this article), the more accurate they will be at determining sentiment and automatically extracting interesting, relevant topics.

In this sense machines really function just like humans; for certain subjects there is a degree of knowledge required to make judgments about conversations and mentions. To demonstrate, here’s an excerpt from another article by our head of NLP Research, Dr. Taras Zagibalov:

“It has grown by 10%”

Is this good or bad? Firstly of course, the answer depends entirely on what “it” is (for instance, income or unemployment) and secondly what we know about growth in that context; is 10% a good or bad amount to grow by? Is growth a good thing at all? Ambiguities like this are not rare; it is extremely common that, to be analysed accurately, pieces of text require some expertise or knowledge that is not commonly possessed.

“The delivery was good”

An academic study showed that, in the context of eBay user feedback, the word ‘good’ is in fact a slight indicator of negativity. Someone without much online selling experience may conclude that the above is positive while the same review may upset a seasoned eBay seller. Similarly, for an ultra-luxury brand ‘good’ might not be good enough.

“The price has dropped, it’s really cheap now”

A final example to illustrate the perspective-dependent nature of any sentiment analysis – the above may be good news for those interested in buying the product, but shareholders of the company selling it will be less pleased about the implications of the statement.

The fact that the machines may lack this knowledge is a deficiency in preparation not in the technique itself.

Tools built purposefully for one industry

With the linguistic challenges in mind, is it unrealistic for one tool to cover multiple, contrasting industries?

Well, whilst a tool that specializes in one specific industry may well offer stronger language analysis, competing on other key attributes that benefit from economies of scale will be difficult:

  • The quality and performance of user interface
  • Development of new features
  • Customer service
  • System speeds
  • General understanding of the market and consumer demand

Cruvee’s success does signpost the importance for monitoring tools to recognize the discrepancies between industries and the language found in them.

But, if users continue to demand more bespoke services for their specific industry, perhaps a better direction will be for the established, premium monitoring tools to partner with smaller companies offering industry-specific expertise and work to tailor the product so that the benefits of both can be combined and exploited.