Deloitte published an article on this subject in 2014 entitled Cognitive Analytics. The article lists three key elements of Cognitive Analytics:
In practical terms, cognitive analytics is an extension of cognitive computing, which is made up of three main components: machine learning, natural language processing, and advancements in the enabling infrastructure.
I get the machine learning (ML) part. The definition I prefer for ML is this:
The field of CS and Math that is related to algorithms and programs that can improve their performance through experience.
Experience means more data, more training time, more examples, etc.
If we are talking about adding “natural language processing” (NLP) to whatever the other analytics in the slide above can do, then I like it. Natural language processing goes beyond speech recognition and allows to extract the meaning out of words, sentences, texts, and unstructured data. So, this should grant access to more data that is in various unstructured formats:
NLP allows more data from more sources to be included in an analysis—allowing raw text, handwritten content, email, blog posts, mobile and sensor data, voice transcriptions, and more to be included as part of the learning. This is essential, especially because the volume of unstructured data is growing by 62 percent each year and is expected to reach nine times the volume of structured data by 2020. (source: Deloitte)
NLP is an interesting field of science on its own, and, when most of the research objectives of this field are achieved, our world will change noticeably. And I am talking about going far beyond Siri or even Watson.
However, this combination of ML and NLP doesn’t justify, IMHO, the “coining” of a new field of science and calling it Cognitive Analytics.
And the role of the right infrastructure to enable the above two technologies to work fast and at low cost is clearly critical but, still, there is nothing novel to this discussion that wasn’t covered in such a field as, say, Big Data Analytics. And Watson is not even a true supercomputer. Maybe, the reference above is to those computers that are optimized for AI/ML tasks? Like this one, for example?
So far, when answering the question of whether this following equation is correct, I would vote as “no.”
ML + NLP + Right Infrastructure = Cognitive Analytics
Rather, Cognitive analytics looks like a field of research at the intercept of several other well-established fields of research that is still looking for its identity and a new, distinctive name.
BTW, if “cognitive” is simply a cute marketing term, then I like it. But I also call some of our own Alchemy IoT algorithms “cognitive” because of the way they acquire knowledge, learn, make mistakes, interact with people, and get better over time in the same way we, humans, do.
If this is a reference to Watson and how it “thinks” (read more here), then, again, it might not be enough to qualify as a new field, because the fundamental algorithms inside of Watson are similar to what most ML and AI scientists or statisticians are using.
For further reference, Cognitive Analytics fits well under the umbrella of Cognitive Computing, which, as of today, includes the following characteristics:
- Natural Language Processing
- Machine Learning
- Algorithms that learn and adapt
- Vision-based sensing and image recognition
- Spatial and contextual awareness
- Reasoning and decision automation
- Sophisticated pattern recognition
- Neural Networks
- Semantic Understanding
- Noise Filtering
- Common Sense
- Robotic Control
- Emotional Intelligence
You can see the overlap here…
P.S. In the next few years, we might see a lot of new fields using the word “cognitive.” In many cases, this will be like the addition of “.com” in 1999 to the name of otherwise traditional companies after taking them online. But, in some cases, this will represent a true and powerful change towards smarter ways to interact with and use computers. The real evolution of analytics.