AI key terms

• Information technology — Artificial intelligence — Artificial intelligence concepts and terminology
• Technologies de l'information — Intelligence artificielle — Concepts et terminologie relatifs à l'intelligence artificielle
• INTERNATIONAL STANDARD ISO/IEC 22989 First edition 2022-07
• Terms related to AI

  • Terms related to data
  • Terms related to machine learning
  • Terms related to neural networks
  • Terms related to trustworthiness
  • Terms related to natural language processing
  • Terms related to computer vision
• Strong and weak AI


  General and narrow AI
• Cognition


  The mental action or process of acquiring knowledge and understanding through thought, experience, and the senses
• Cognitive computing


  Computing systems that are inspired by the human brain and nervous system
• Semantic computing


  Computing that is focused on the meaning of information
• Soft computing


  An approach to computing that focuses on approximate solutions to computationally hard tasks
• Genetic algorithms


  Optimization algorithms inspired by the process of natural selection
• Symbolic and subsymbolic approaches for AI

  Symbolic approaches use explicit representations of knowledge, while subsymbolic approaches use distributed representations
• Supervised machine learning


  Machine learning where the training data includes the desired outputs
• Unsupervised machine learning


  Machine learning where the training data does not include the desired outputs
• Semi-supervised machine learning


  Machine learning that uses a combination of labeled and unlabeled data
• Reinforcement learning


  Machine learning where an agent learns by interacting with an environment and receiving rewards or penalties
• Transfer learning


  Applying knowledge gained from one task to a different but related task
• Training data


  The data used to train a machine learning model
• Trained model


  A machine learning model that has been trained on data
• Validation and test data

  Data used to evaluate the performance of a trained machine learning model
• Retraining


  The process of updating a trained machine learning model with new data
• Examples of machine learning algorithms

  • Neural networks
  • Bayesian networks
  • Decision trees
  • Support vector machine
• Autonomy


  The ability of a system to operate independently without external control
• Heteronomy


  The state of being under the control or influence of an external source
• Automation


  The use of technology to perform tasks without human intervention
• Internet of things


  The network of physical objects that are embedded with sensors, software, and connectivity to enable these objects to collect and exchange data
• Cyber-physical systems


  Systems that integrate computation, networking, and physical processes
• Aspects of AI trustworthiness

  • Robustness
  • Reliability
  • Resilience
  • Controllability
  • Explainability
  • Predictability
  • Transparency
  • Bias and fairness
• AI verification and validation


  The process of ensuring that an AI system meets its requirements and performs as intended
• Jurisdictional issues


  Legal and regulatory considerations related to the deployment of AI systems
• Societal impact


  The effects of AI systems on individuals, communities, and society as a whole
• AI stakeholder roles


  • AI provider
  • AI producer
  • AI customer
  • AI partner
  • AI subject
  • Relevant authorities
• AI system life cycle

  1. Inception

  2. Design and development

  3. Verification and Validation

  4. Deployment

  5. Operation and monitoring

  6. Continuous validation

  7. Re-evaluation

  8. Retirement
• Data and information


  The inputs and outputs of an AI system
• Knowledge and learning


  The internal representations and processes that allow an AI system to acquire and apply knowledge
• From predictions to actions

  1. Prediction

  2. Decision

  3. Action
• AI ecosystem


  The interconnected network of AI systems, stakeholders, and supporting infrastructure
• AI ecosystem


  Interconnected entities, people, systems and information resources together with services that process and react to information from the physical world and virtual world
• AI systems


  Engineered systems that generate outputs such as content, forecasts, recommendations or decisions for a given set of human-defined objectives
• AI function


  The purpose or role of an AI system
• Machine learning


  A category of AI techniques that enable systems to learn and improve from experience without being explicitly programmed