What is Cognitive Computing?

Cognitive computing is one of the core technology sub-fields of cognitive science, an important part of artificial intelligence, and a computer system that simulates the cognitive process of the human brain. ^[1]

Cognitive computing

The simplest tasks of cognitive computing are speaking, listening, seeing, and writing, and complex tasks are assisting, understanding, decision-making, and discovery. Cognitive computing is a top-down, global, unified theoretical study designed to explain The observed cognitive phenomena (thinking) conform to the known bottom-up neurobiological facts (brain), which can be calculated or explained using mathematical principles. It seeks a kind of software and hardware components that conform to the known computer sciences with the basis of brain neurobiology, and is used to deal with mental processes such as perception, memory, language, intelligence and consciousness. One goal of cognitive computing is to enable computer systems to learn, think, and make the right decisions like the human brain. The human brain and computer each have their own strengths. Cognitive computing systems can become a good auxiliary tool to cooperate with humans to solve some problems that the human brain is not good at. ^[1]

Cognitive computing has a very important impact on the fields of artificial intelligence, information technology, and cognitive science in the future.

Cognitive science dates back to the 1950s, and the name was proposed in 1956 at a Massachusetts Institute of Technology scientific seminar on information theory. In the 1960s, cognitive science began to develop. In 1976,

MIT economist David Autor analyzes more than 100 applications of cognitive technology and divides these applications into three major categories: product categories,

Process and analysis classes. Each type of application has a profound impact on work and workers.

Product applications embed cognitive technology into products to enable intelligent behavior, natural communication (such as seeing and seeing), and automation. The impact of such applications on workers ranges from no impact (robot toys or smart thermostats) to some impacts (sweeping robots that can handle household cleaning tasks) to significant impacts: autonomous vehicles are replacing mining truck drivers and train drivers, One day they will replace the work of taxi or truck drivers; robots will also replace masons and bricklayers. In the next step, companies can deploy products using cognitive technology into various business processes.

Process applications use cognitive technology to enhance, expand, or automate business processes. Examples include automated data entry, automated handwriting recognition, autonomous planning and adjustment using algorithms, and automated customer service using speech recognition, natural language processing, and question answering technologies. By defining process applications, workers' work can be fully or semi-automatic.

Analytical applications use cognitive technology to uncover patterns, make predictions, and guide more effective actions. For example, Intel uses machine learning to show its sales teams further customer needs and what they should offer customers. Some analytical applications can be seen as a form of automation: what happens next is a decision made by a machine, not a person, based on a particular situation.

At present, many technology companies have begun to use cognitive computing to improve their business level, or to help other companies transform, creating more and more cases of cognitive business and cognitive enterprise success.

Intel

Intel Corp. has developed a cognitive system to increase sales. The system uses machine learning to classify customers and instruct sales staff on which products to sell to different customers.

Intel's RealSense technology platform. The RealSense technology development platform and its software development tools (SDK) were released in 2014, providing developers with tools and resources for developing applications through cognitive computing technology.

Cisco

The Cisco platform called Cognitive Threat Analysis is a cloud-based solution that reduces the time to discover potential threats to the network. The need for cybersecurity threat analysis through behavioral analysis and anomaly detection is increasing. One requirement is that it can identify abnormal behaviors of defenses outside the network, and then identify malware intrusions or data loss to assist behavior analysis and anomaly detection.

IBM

· P53 is an important protein associated with many cancers, and so far 70,000 papers have been published on this protein. The Baylor College of Medicine said that even if scientists read five papers a day, it would take 38 years to fully understand the protein. However, through the collaboration between Baylor College of Medicine and IBM, in a matter of weeks, biologists and data scientists used the Baylor Knowledge Integration Toolkit (KnIT) to accurately identify the modifiable P53 based on Watson technology. Protein ultimately improves the effectiveness of drugs and other therapies. This automated analysis led cancer researchers at Baylor College of Medicine to identify seven potential proteins as targets for new research. In the past three decades, scientists have only made a similar target protein discovery every year.

· IBM and SoftBank Robotics Holdings (SBRH) jointly launched Pepper, an intelligent robot based on Watson CCP. It can communicate with humans normally, can recognize text, images and voices. Through industry customization, it can be used at bank desks, restaurants, retail , Hotels, medical reception and other fields to provide intelligent information services for human beings. ^[4-5]