There is no concrete definition assigned to cognitive computing, but it is broadly described as technology platforms which are based on Artificial Intelligence and Signal Processing. These platforms, billed as one of the key technologies for big data, can encompass machine learning, reasoning, speech, and language processing.
The term cognitive computing is usually used to describe hardware or software that acts or processes information in a similar way to how the human brain functions. This mimicking of the human brain has led to cognitive computing being recognised as a new type of computing, with IBM describing cognitive computing as ‘systems that learn at scale, reason with purpose and interact with humans naturally.’
The ultimate goal of cognitive computing systems is to create automated IT systems capable of solving problems without human supervision or assistance.
It is common for people to think of Artificial Intelligence and cognitive computing as one in the same, but cognitive systems differ in the fact that they comprise of separate components, from different disciplines, working together.
Cognitive computing platforms can consist of many different features depending on the application, but common features can grouped under the labels adaptive, interactive, stateful and contextual.
An adaptive feature of a cognitive system encompasses what is effectively machine learning – the system reacts to changing information and tailors goals and requirements to that information.
Interactive features of a cognitive system denote the ability to easily interact with users in regards to what they need. This is not restricted to just people, with processors, devices and cloud services also able to interact with the system.
Stateful features of a cognitive system is all about problem solving, as well as ‘remembering’. Through asking questions or finding additional information, cogfnitive systems can help in defining a problem. The system can then draw upon past interactions with the user and give them suitable information.
Contextual features of a cognitive system needs little explaining – it is all about giving context to the user. A system could be asked to identify, understand and extract information such as syntax, time, date, task or meaning . The system could draw on structured and unstructured data as well as sensory inputs in order to offer context.
One of the best known cognitive systems is IBM Watson – described by Big Blue as a ‘technology platform that uses natural language processing and machine learning to reveal insights from large amounts of unstructured data.’