Cognitive Fitment

What is Cognitive Fitment

When factor of problem representation and problem-solving tools/methods match the characteristics of the task, we call it Cognitive Fit and the method or study to reduce the characteristic dissimilarity is called as Cognitive Fitment Analysis. Even though we have the word called Cognitive Fitment, that doesn’t necessary mean we are referring to any Automation or Robotics mechanism to undergo the study based on the input and desired output, although we should be have a definite aim to introduce Automation(AI/ML) and Robotics as much as we can. This can be done through all the historical data analysis gathered from manual exercise of Cognitive Fitment.

The good thing about Cognitive Fitment is it can touch any dimension and paradigm of customer business and accompany them to realize the performance accrue through this exercise, like it can touch physical process of “order-pickup-deliver” mechanism and it can also touch “technical fitment analysis of an application to run on certain infrastructure”

The Cognitive Fitment exercise will also help the organizations to understand the Mental Representation of the Customer. To understand, if the customer itself is inspired to solve the problem, if he is making a correct representation and he is clear about his end goal is a major milestone reached in terms of shaping up the business which includes target derivatives.

The more Cognitive Fitment study any organization does, for its numerous customers, it grows its own Knowledge Database about it, which can be easily re-utilize for similar brands or similar supply chain mechanism.

To create Cognitive Fitment knowledge DB, the organization has to deep dive to understand what are the tools we have in today’s market as an offering to utilize to make the study or if we really have a tool, or it needs to be internally developed based on seller experience and how one can initially start the manual exercise.

Cognitive fitment will touch the component of “Customer Delight”, “Unified Customer Experience” and “Omnichannel Collaboration”.

Why Cognitive Fitment

We stress a lot these days about Cognitive abilities, be it a person or be it an intelligence workspace, business and so on. Cognitive skills help your brain complete information process more quickly and efficiently, and you ensure that you understand and effectively process that new information. Your cognitive abilities help you process new information by taking that information and distributing it into the appropriate areas in your brain. When you need that information later, your brain also uses cognitive skills to retrieve and use that information. By developing cognitive skills, you help your brain complete this process more quickly and efficiently, and you ensure that you understand and effectively process that new information.

Quite similar to that, day to day business needs intelligent data processing, distribution and decision making based on the need of the problem or subject, to stay ahead in the game than its competitors. If we consider entire business of the organization is a big data pool where every action and reaction have a meaning, then there is definite pattern for the said meanings. Cognitive Fitment is a analytical ability to derive pattern based on situation and appropriately store, process and re-validate them based on identical situation.

Just think of situation, you are standing in Burger shop and seller knows from your face reaction, you want to eat a special variant of Chicken Burger. You will be surprised isn’t it? And more than that, you come back again to the same shop, other than going around to other Burger shops, to test the seller again or the process, if it gives you the same result. You may be quite intrigued to find out, this time, there is a different seller on the shop and yet he identifies, what you desire to eat. Astonishing isn’t it? You may be perplexed what and how the shop is running the game or the sellers, but what you don’t know, it’s the Cognitive Fitment study done over a period of time, with massive amount data collection of face pattern data, which has lead to them to your(customer) delight.

Now this is something which has a significant similarity between human interpretation and machine interpretation within an added advantage for machine interpretation as it can factor out molecular details, which human eye has a great possibility to overlook. When we talk about similarity, lets consider of your workplace. Imagine you are working in the same place under the same supervisor for a good amount of time, where you have developed an idea what will make your supervisor happy and vice versa. So a good fated positive supervisor may have already noticed what you are good at, what kind of work brings the best out of you, and which work may be your dislike but still you manage to deliver it on ordinary standards. And here he can balance it out properly, which brings the best out of you for the organization. And all these is because he has observed you and your working style for a significant amount of time. Now imagine’ your bosses’ position when a newcomer joins the team. The supervisor does not a great deal of data about the person, so how will he judge him? He will judge him based on the similarities and dissimilarities of people whom he has already observed for in his career for a certain period. He will have several style buckets in his brain which will stimulate to create an initial image about the person based on his matching and mismatching traits from those brain buckets, which may change later or may not based on other different traits the new comer exhibits, which has not exposed earlier, or he may come out a completely different trait which doesn’t match to anything earlier, which was there in your bosses’ brain. Now we may call that as an “Exception” and every one of us may agree as a matter of fact, every subject or field study does have exceptions to deal with.

Now if we have to develop similar interpretation models like the way our brain functions, we need to have a whole lot of pattern data(read as traits) to be developed in the pool, for artificial intelligence to continuously read them with Machine Learning algorithms and derive the benefit outcome of it, which is what we called as “Cognitive Fitment” and it may be need of the hour to elevate your business to next level, by understanding customer traits to bring them much delight. If your customer is successful through your efforts, then you are successful.

How we can do Cognitive Fitment?

We can have multiple ways to do it. Let’s get back to our previous example of the Burger Seller. The Burger could have done the little magic in 2 different ways:

§ Use Case-1: The customer(you) may have been a regular visitor to this Burger shops or outlets. It may be one outlet or multiple other outlets which basically doesn’t make any difference as the data whether stored centrally or in a distributed fashion will be a part of the algorithmic analysis overall. The most important thing is what exactly he has done behind the scenes with the data. The burger organization must have been running “A Facial Expression Recognition Algorithm Based on Feature Analysis” which has inculcated itself with specific patterns based on the timing, weather, day, month, week of the month, weight time in the queue etc. Along with the same another algorithm has also learnt the customer’s eating patterns. With the both the feed, the parent algorithm has developed a fare notion about the customer’s body language while he/she approaches the counter, monitored with high definition cameras with integrated AR (Augmented Reality) Glasses.

§ Use Case-2: The second use case is based on the case, where the customer is visiting the Burger shop for the very first time. A tough call for the algorithm but not impossible. Like the above example about the supervisor, the algorithm do have the chance to compare the feature extracts and match the equivalent features within its database and come up with the recommended action based on the same database which it has collected and ordered over a period of time. This is quite similar to the way how “Biometric matching of fingerprints or other components are done on the Forensics”. Now, as I told earlier, there can be exceptions. It may well happen, the feature extract of the new customer, may not have any close match with what is there in feature database, which may well happen at the very beginning during the initial days of the implementation quiet frequently and that is expected. Nothing to be disheartened about. The application may help the seller with the instruction — “New customer, new territory to explore and learn 😊”

A small note on Facial Expression Recognition Algorithm:

A great amount of research on this specific area has shown multiple different ways to achieve the outcome.

One is with “Feature Fusion” where the algorithm, initially segments the eyebrows, eyes and mouth areas from the facial expression images, and computes it with Auto-Correlation method, with reduced dimensions through Weighted Component Analysis. Followed by that, the weights values are obtained according to facial expression measure system and finally minimum-distance classifier will recognize different expressions.

Another way is to do “Geometrical feature extraction based on emotional expression classification using machine learning algorithms” which defines “virtual markers” on the face of the subject and forms automated triangles. The movement of the markers during facial expression directly changes the property of each triangles. The area of the triangle, circle circumference, and the circle area of a triangle are extracted as features to classify the facial emotions. These features are used to distinguish facial emotions using various types of machine learning algorithms. The maximum mean classification rate of the circle area of the triangle will determine the best fitted feature of the geometrical analysis of facial expression and it is often obtained by Random Forest (RF) classifier near to accuracy, compared to other classifiers in the game.

Now as our subject is not around facial recognition, we will not go any further on it, on this article.

When was Cognitive Fit Theory developed?

Cognitive fit theory was developed by Iris Vessey (1991). The theory proposes that the correspondence between task and information presentation format leads to superior task performance for individual users. In several studies, cognitive fit theory has provided an explanation for performance differences among users across different presentation formats such as tables, graphs, and schematic faces (e.g., Vessey, 1991, 1994; Vessey & Galletta, 1991; Umanath & Vessey, 1994). The theory has also been extended into the geographic information systems domain, where it has been used to explain performance differences among users of map and table-based geographic information systems on adjacency, proximity, and containment tasks (Dennis and Carte,1998; Smelcer and Carmel, 1997). [ Gathered from theorizeit organization].

Cognitive Fitment Platform — Sample High Level Flow Infusion


Decision support systems to satisfy business desires have a large effect both efficiency and effectiveness of problem solving through Cognitive Fitment study. Determining the characteristics in name of features of the subject and approaching the best fit solution gathered from the recommendation of Cognitive Fitment study can change world for ever in terms of on the job, ubiquitous marketing. Enough evidence now exists to suggest that the notion of cognitive fit may be one aspect of a general theory of problem solving. Suggestions are made for extending the notion of fit to more complex problem-solving environments which may cover Expression Recognition, Business Pattern recognition, merger analysis, weather treatment, medical symptoms analysis etc.

Thanks for reading my article on Cognitive Fitment. Please 👏if you have liked it and let me know your feedback on the same. In case you want to read more about the subject, do let me know. I’m avid researcher on this and can come up with more detailed versions of it.




Bijoy is distinguished IT Enterprise Architect for Multi Cloud(Azure, AWS, GCP), Datacenter, DevOps, Automation and SRE technologies, having 15.5 years of exp.

Love podcasts or audiobooks? Learn on the go with our new app.

When AI meets Biology

Today’s Supercomputers will be tomorrow’s wristwatches — processing power can be the key to unlock…

supercomuter to wristwatch

The Hidden Treasures in Language

Artificial Intelligence — 10 fears answered — The Digital Transformation People

Intelligence as a source for Innovation

Tsinghua AI Association of International Students (TAIS) news roundup 18 May

The Effect DAO: The Constitution

Introduction to Reinforcement Learning

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Bijoyendra Roychowdhury

Bijoyendra Roychowdhury

Bijoy is distinguished IT Enterprise Architect for Multi Cloud(Azure, AWS, GCP), Datacenter, DevOps, Automation and SRE technologies, having 15.5 years of exp.

More from Medium

People’s participation in Nepal amid COVID 19 pandemic

The Culture Map: How culture shape our ways of thinking?

Trailblazing Data-Driven Narrative Change

Do you think of yourself as intelligent and sophisticated?