Answer engine
A system that responds to a question with a direct answer rather than only returning a ranked list of links. Some answer engines retrieve current sources before generating a response.
Plain-language reference
AEO Glossary
Definitions for the technologies, practices, and measurements used throughout AEOsrc.
Answer Engine Optimization (AEO)
The practice of making information easier for answer-producing systems to discover, understand, retrieve, and use. AEO builds on SEO rather than replacing it.
Citation
A visible reference connecting part of an AI-generated answer to a supporting source. Citation interfaces and source-selection methods differ by platform.
Context window
The amount of information a language model can consider during a single interaction. Retrieved passages, instructions, and conversation history may all consume part of this space.
Crawler
Automated software that discovers and requests web content. Search engines and some AI companies operate crawlers for different purposes and under different user-agent names.
Embedding
A numerical representation of content that allows software to compare meaning or characteristics mathematically. Embeddings are commonly used in semantic retrieval and recommendation systems.
Entity
A distinct person, organization, product, place, event, or concept that a system can identify and connect to attributes and relationships.
Generative search
A search experience that uses a generative model to synthesize information into a response, often with links or citations to supporting sources.
Grounding
Connecting a generated response to external evidence, such as web pages, documents, databases, or search results, to improve freshness and verifiability.
Index
An organized collection of processed information designed for efficient retrieval. Search indexes may store terms, metadata, links, vectors, or other representations.
Knowledge graph
A structured representation of entities and the relationships among them. Knowledge graphs can support disambiguation, discovery, and questions involving connected facts.
Language model
A model trained to process and generate language by learning statistical patterns from large collections of text and other data.
Lexical search
Retrieval based primarily on words, phrases, and token overlap. Modern systems often combine lexical and semantic methods.
Reranking
A later retrieval stage that reorders an initial set of candidates using a more precise relevance model or additional signals.
Retrieval
The process of finding documents, passages, records, or other evidence relevant to a question or task.
Retrieval-Augmented Generation (RAG)
An approach that retrieves external information and places it into a model’s working context before the model generates an answer.
Semantic search
Retrieval that attempts to match meaning and intent, not only exact keyword overlap.
Source attribution
The practice of identifying where information came from. Attribution may appear as inline citations, footnotes, supporting links, or source panels.
Structured data
Machine-readable markup that describes the meaning and properties of page content using a defined vocabulary, commonly Schema.org.
Vector
An ordered list of numbers. In AI retrieval, a vector often represents an embedding for text, an image, a product, or another item.
Vector database
A system designed to store vectors and retrieve the items whose vectors are closest to a query vector, often alongside metadata and original content.