Unlocking Internet intelligence with text analytics

How such a ‘dull’ discipline as text analytics can become the ‘hot stuff’ when stimulating the money stream.

Here we are. It’s the XXI century - an era of unlimited opportunities propelled by a global network that makes everything easier, faster, and available directly at our fingertips - a global marketplace, business without borders, international partnerships, customers everywhere.

Is it really so?

The growth of the Internet unveiled greater opportunities, and at the same time flooded us with incomprehensively large volumes of data, making it much more difficult to effectively extract useful information from the vast number of online sources.

Actually, it's quite difficult to locate potential opportunities when we try to target the whole world, and without a better way of aggregating and organizing vast and disparate caches of information, we won't avoid drowning in digital noise.

The Web shrouds a real gold mine

A significant amount of valuable information is carried by unstructured data, which constitutes 80-90% of the world's data resources.

For business data is a competitive currency. The main source of relevant information originates in unstructured text format, and as such, the web offers a large collection of this demanded currency.

We already have many data nuggets in our possession - emails, messages, documents. Yet, the Web holds millions of websites, publications and offers, text generated by people’s interactions, and topical search queries, which all could, if cleverly approached, give us new insight about customers' needs, profitable business partnerships, and new revenue opportunities.

An ability to achieve this Internet intelligence will give us access to a real gold mine.

Dear stalking ads – I don’t need you anymore

Beginning our business seeking with ready-to-use matched and customized information, rather than sifting through piles of digital rubble, and being able to immediately interact with customers who are starting their search for a solution or product, will affect many industries and greatly simplify some important business processes.

We could stop relying mainly on tracking users across the web and foretelling their behaviour to push them our ads which, despite of claiming that they are intent-based, have seldom anything to do with the searcher’s true intention. They are ever more viewed upon as invasive, annoying, even stalking, and so result in poorer business image. We often close or block them with no regret (as a marketer, do you really content with your ROI?).

Instead, we could change this approach and start using real intent-based solutions that serve information to an actual matched need, meet demand requirements in real-time, and are truly welcomed by both sides - whether a customer or a seller.

The rising power of text analytics

Unleashing the potential of this data, and bringing together diffused content and untapped opportunities, is a vast and extremely tempting task that should be done by machines, rather than relying on humans alone.

We put great hope in progress of AI, machine learning (ML), and related technologies, which make it possible to do things that would be impossible otherwise.

The key to accessing unlimited Internet sources is the ability to understand unstructured data, and the key to understanding this data is text analytics.

Text analytics (roughly equivalent to text mining, which is a more scientific term) isn't about sentiment analysis claiming to give real insight on what customers think (fortunately this hype has lately calmed down). It also isn't about keyword indexing, and not even about key phrases extraction.

Text analytics is about sense, meaning, and the ability to organize and classify information, making it useful to people. It's also essentially our dream gateway to general artificial intelligence (AGI), where "thinking machines" (that's how we define/imagine it now) truly understand and perform intellectual human tasks, not only use programmed tricks to mimic behaviour.

Money talks

It's not that we couldn't somehow analyze and use unstructured text data before. We just didn't do it correctly enough, and not at a rewarding scale yet. Need simple proof instead of extensive scientific explanations? Follow the money. You will see a score of technology giants and Venture Capital companies (VCs) investing in text analytics startups, as well as top technology research reports arguing that the rise of text analytics and its market application is still ahead of us.

This promising discipline, as part of artificial intelligence (AI), involves many related fields and methodologies, like data mining, information extraction, lexical analysis, and natural language processing (NLP).

And no matter how ordinary and non-AI it might seem, it's probably one of most important technologies with potential to impact and accelerate market change on a global scale... and it's just at the very beginning of its growth trend.

Flooded with opportunities

Text analytics-based software able to autonomously classify web content with reason and understanding, can be used to build large collections of ready-to-use information, which streamlined on demand to the right people at the right time could make our digital business experience far more effective.

Automated gathering and processing of unstructured data is necessary to make Internet intelligence extraction real and scalable.

After achieving that, we can start thinking about a digital intent-based business ecosystem powered by a self-learning solution providing real-time customized answers to actual demands.

This is where machine learning, maybe even deep learning (so fashionable recently) when it finally matures, can begin.

Collecting information to be processed by machine learning applications for later on-the-fly automated reasoning seems to be the logical path, especially that machine learning is math, and as such it requires data which fuels learning.

Combining text analytics, large scale data gathering, and unsupervised information processing with machine learning could serve businesses with a real AI-based demand-generation supporting technology, positively impacting many industries.

We could finally stop wasting time searching and focus on what really matters to us, staying informed of new opportunities and putting our efforts to relation nurturing and making business.


You may be also interested in: Look beyond the query