Marketing & Advertising Glossary

Rate limiting controls how many requests an application can make to an API within a specific time period to prevent overload.

An AI system that predicts and suggests products, content, or ads most likely to interest individual users based on their behaviour and preferences.

A machine learning technique where AI models are trained using human feedback to improve accuracy, safety, and relevance in marketing applications.

Responsible AI refers to the ethical development and deployment of artificial intelligence systems that are transparent, fair, and accountable in advertising and media buying.

Software automation that mimics human actions to execute repetitive marketing and media buying tasks with minimal human intervention.

Semantic search interprets the meaning and intent behind search queries rather than just matching keywords, using AI and natural language processing.

Sentiment analysis uses AI to detect and classify emotions and opinions expressed in text, helping brands understand audience reactions to campaigns.

Unauthorised or untracked AI systems operating within marketing and media workflows without proper governance or oversight.

Stable Diffusion is an open-source AI model that generates high-quality images from text descriptions, enabling marketers to create visuals without hiring designers.

Two fundamental machine learning approaches: supervised learning uses labeled data to predict outcomes, while unsupervised learning finds hidden patterns in unlabeled data.

Artificially created data generated by machine learning models to simulate real-world patterns without exposing sensitive information.

Temperature controls randomness in AI model outputs, affecting creativity and consistency in generated text and predictions.

Throughput measures how much data or how many requests an AI system can process within a given time period, typically per second.

The historical information used to teach AI models how to recognize patterns and make predictions for advertising and marketing decisions.

Transfer learning applies knowledge from pre-trained AI models to new advertising tasks, reducing training time and improving performance with limited data.

A neural network architecture that uses attention mechanisms to process data sequentially, powering modern AI systems like ChatGPT and predictive advertising tools.

Weights are numerical values in machine learning models that determine how strongly input data influences predictions and outcomes.

Streaming in AI advertising refers to the continuous flow of real-time data used to train models, make predictions, and optimize campaigns instantly.

DALL-E is an AI image generation tool that creates original images from text descriptions, useful for marketers creating ad creative and visual content.

Deep learning is an AI technique using multi-layered neural networks to identify patterns in data and make predictions without explicit programming.

Inference is when a trained AI model makes predictions or decisions on new data in real-time, without learning from it.

Kubernetes is an open-source platform that automates the deployment, scaling, and management of containerized applications.

Perplexity AI is an answer engine using large language models to provide sourced, conversational responses to user queries.

RAG is an AI technique that combines document retrieval with language models to generate accurate, contextual responses using real-time data.

A TPU is a specialized AI chip designed to accelerate machine learning tasks, increasingly used in ad tech and programmatic advertising platforms.

A chatbot is an AI-powered software that simulates human conversation to interact with users, answer questions, and automate customer service.

An AI agent is a software system that perceives its environment and takes independent actions to achieve specific goals.

An epoch is one complete pass through an entire training dataset during machine learning model development.

Using AI and software to automatically execute repetitive marketing and advertising tasks, saving time and improving consistency.