What is Artificial Intelligence and what are cognitive sciences?
Intelligence is a human thing!
Intelligence can be defined as one’s capacity for logic, understanding, planning, reasoning, making decisions and problem solving.
It is the ability to perceive information and to retain it as knowledge to be applied towards adaptive behaviors within an environment or context.
And this is what cognitive scientist do. They try to understand how intelligent processes are working. Cognitive science is the interdisciplinary, scientific study of the mind and its processes and cognition is the mental action of acquiring knowledge and understanding.
And this is what we do at YellowOrange. We put intelligence inside something immaterial, into something artificial called a software. In short, what we do is Artificial Intelligence (AI).
And how to build an Artificial Intelligence?
The first thing to understand when you want to build an artificial intelligence is what mental processes are. These are well known and can be grouped in those used to perceive and learn in order to categorize knowledge and those who will use this knowledge to make something intelligent: reasoning, making decisions and solving problems.
And for cognitive processes we have AI solutions: these are for reasoning, fuzzy logic, decision making, problem solving and categorization
Based on these solutions, we want to share with you some of our success stories in this area:
Hector : word comparison algorithm based on fuzzy logic and to be used within data mining / big data solutions. This algorithm was embedded within a virtual personal assistant called Hector and won a national innovation award (intelligent text search engine) in 1994.
Laforian / INRETS : expert system who had to evaluate the risks taken by a pedestrian while crossing a road (dynamical problem space).
Aile-SYS : real time video analysis software used on highways to detect dangerous situations and send alerts to the patrol. Installed outside, this system took into account weather changes (trees moving, day, night, as well as normal moving cars).
UCB : intelligent visual pallet conveyor system who is managing proactively a storage warehouse for the pharmaceutical industry. The system was able to solve conveying problems and conflicts to accelerate pallet processing.
Société Générale : fuzzy logic and neural network used to identify (machine learning) and manage software assets during a manual inventory of 2400 user workstations. Output used to manage licenses, software packaging and OS migration.
Andersen Consulting : ecological graphical user interfaces study. Cognitive and neurobiological approach to understand how humans decode information while using a software. Ergonomics best practices then used to design user friendly interfaces for clients.
ASML : neural network used to categorize software assets in order to create functional heatmaps (knowledge representation) and make decisions about how to rationalize the software portfolio (1200 different software were used, only 60 remained).
Confidential : deep question and answer software made to simulate the presence of a human during an online chating session to maximize the presence of the customer and to minimize the interactions with a teleoperator (virtual language processing and generation).
CEA : real time monitoring and analysis of atmospheric particulate matter using fuzzy logic to proactively forecast abnormal emission levels before they occur and are released in the environment.