Curiosity

Glossary

Vector/embedding-based retrieval that finds results by semantic similarity rather than exact keywords.

An alternate string representation for a node (e.g., punctuation variants or abbreviations) used to improve matching and linking.

Splitting long text into smaller pieces so embeddings and NLP can process it effectively without losing important content.

A code-driven ingestion component that reads data from external sources and maps it into node/edge schemas in a Curiosity Workspace.

A relationship between two nodes in the graph (e.g., Case -> ForDevice -> Device).

A vector representation of text (or other content) used for semantic similarity and vector search.

A server-side function hosted inside a workspace. Endpoints implement business logic, orchestration, and custom APIs.

A filter dimension presented to users (e.g., Status=Open, Type=SupportCase). Facets can be property-based or derived from graph relationships.

A labeled property graph of nodes and edges representing entities and their relationships.

Retrieval strategy that combines keyword search and vector search to improve both precision and recall.

A data structure built over selected fields to support fast retrieval (text search index and/or embedding index).

A field (or set of fields) that uniquely identifies a node within a node type. Keys enable idempotent ingestion and safe updates.

A configurable sequence of NLP steps applied to a field (tokenization, entity capture, linking, etc.).

An entity in the graph (e.g., User, Ticket, Device) with a type and properties.

A field stored on a node used for display, filtering, and indexing (e.g., Summary, Status, Time).

The definition of the node types and edge types in your workspace, including keys and properties.

An environment that contains your graph data, schema, search indexes, NLP/AI configuration, endpoints, interfaces, and administrative settings.