Glossary
Access levels (Read / Write / Admin) — Workspace permissions that control what you can do. Read lets you view and search. Write adds uploading and editing files, creating and editing records, and managing tags/comments. Admin adds workspace configuration and member/role management.
Access Manager — A Fusion Center page used to review users, roles, and effective permissions, and to onboard or update access.
Administrative workspace — A workspace used to store configuration assets (for example apps, templates, dashboards) rather than day-to-day evidence. It is often configured to Auto open and Hide by default for standard users.
AI Chat — A chat interface that answers questions using a defined set of sources. When you add context (files, folders, records, sets, or workspaces), responses are grounded in that evidence and limited by your permissions.
App — A task-focused tool that runs inside Octostar (for example Entity Extract & Graph, CSV Importer, Transcriber, Map, Report Creator). Apps can take files, records, or sets as input and write results back to the workspace.
App Editor — The in-platform workspace used to view and edit an app’s project files, manage secrets, preview or run the app, and review logs (where enabled).
App Launcher — The menu that lists apps available to you. You can open apps from here or from Open with… on a file or record.
Attachment — A file linked to a record (for example a document, image, transcript, or export) to support provenance and collaboration.
Audit log — A record of user actions and system events used for traceability (availability depends on deployment and permissions).
Cards view — A Search or dataset view that shows results as card tiles with key attributes and quick actions.
Concept — A type of entity defined in the Ontology (for example Person, Organization, Event, Place, Object). Concepts determine which fields (properties) and links (relationships) are available.
Concept mapping — A mapping that turns datasource rows into entities of a concept, binding columns to ontology properties and defining keys/labels.
Context (AI Chat) — The specific sources AI Chat is allowed to use for an answer (for example selected files, folders, records, sets, or workspaces).
CSV Import — A tool that imports a CSV file into Octostar. Depending on configuration, it can create tables, records, or structured entities.
Datasource — An external system or database connected in Fusion Center (for example MySQL, PostgreSQL, ClickHouse, Redshift, Vertica, SAP HANA). Datasources can be exposed through mappings without moving the data.
Dataset view — A visualization of a result set (for example table, cards, or detailed view). A dataset typically comes from Search or a saved set.
Detailed view — A Search or dataset view that shows richer, expanded attributes per result than Cards or Table.
Embeddings (vector search) — Numeric representations of text or media used for semantic search (for example finding “luxury vehicle” images even without those words).
Entity — A specific instance of a concept (for example a particular person or organization). Entities are represented as records.
Entity Extract & Graph — An app that extracts entities and relationships from a document, highlights them in context, and can normalize them into structured records and a Link Chart.
Filtered concept — A concept plus filters (for example “People with age between 30–40”). Filtered concepts drive Search results and dataset views.
Force re-index — A workspace action that refreshes indexing and enrichment outputs for files or folders so they appear correctly in Search (availability depends on configuration).
Fusion Center — The administrative hub for modeling the Ontology, connecting datasources, creating mappings, managing jobs and access, and validating data (for example via SQL Lab).
Hierarchy level — A concept’s position in the ontology inheritance tree. It helps explain which properties are inherited from parent concepts.
Knowledge graph — A graph representation of entities and relationships defined by an Ontology. In Octostar it is often virtual, with data queried in place.
Label pattern — The rule that defines how an entity is displayed in the UI (for example combining name + date of birth).
Link Chart — An interactive network view that shows entities as nodes and relationships as links. You can expand, filter, layout, annotate, save, share, and export charts.
Many-to-many mapping (relationship mapping) — A mapping that produces relationships (edges) between entities, typically using join logic across tables/views.
Member Access — The workspace dialog where you grant access to users/roles and configure options such as Auto open and Hide by default (depending on permissions).
Metadata (file) — System and extracted attributes about a file (for example type, size, timestamps, EXIF/geo, tags) used for filtering and provenance.
Multi-value mapping — A mapping that expands lists/arrays into repeatable properties or one-to-many values.
Ontology — Octostar’s data model. It defines concepts, properties, and relationships and ensures consistent behavior across Search, records, Link Charts, apps, and dashboards.
Ontology Explorer — A Fusion Center tool for browsing concepts, properties, and relationships, including inheritance and definition details.
Ontological CSV Importer — A CSV import workflow that maps columns directly to ontology concepts/properties/relationships to create structured entities.
Open with… — A menu that opens a selected file or item in a specific app (for example “Open with → Entity Extract & Graph”).
Permissions — Rules that determine what a user can see or do. Permissions apply across workspaces, files, records, and actions, and are usually managed via roles and access levels.
Primary key — The identifier used to uniquely represent an entity in a mapping or concept definition (critical for deduplication and updates).
Provenance — Traceability back to source evidence. It links structured facts (records/relationships) to the underlying documents, files, or datasource rows.
RAG (retrieval-augmented generation) — An AI pattern where the system retrieves relevant context from your selected evidence before answering, improving grounding and reducing guesswork.
Record — A structured item representing an entity (an instance of a concept). Records store properties and relationships and open in the Record Viewer.
Record template — A configured layout for displaying or editing records of a concept (for example a profile view or timeline), where supported.
Record Viewer — The UI used to view and edit a record’s fields, relationships, attachments, comments, and tags (based on permission).
Relationship — A typed connection between two entities (for example works_at, located_at, communicates_with). Relationships may include attributes (for example dates, roles, or confidence) depending on the model.
Saved search — A stored Search definition you can reuse or share (where enabled). Saved searches often produce a set that can be visualized or sent to apps.
Schema Editor — An admin tool used to control how ontology properties appear in forms and Search filters (for example field order, validation, visibility).
Search (Global) — Concept-aware search across unstructured evidence and structured entities. Supports filtering, multiple result views, and saving queries.
Semantic search — Search that uses embeddings to retrieve results by meaning rather than exact terms, across text and media.
Set — A collection of items (often records, sometimes files) produced by Search or selections. Sets can be visualized, saved, and used as inputs to apps.
SQL Lab — A Fusion Center tool for ad-hoc SQL queries against datasources (where enabled) for validation and exploration.
Table view — A Search or dataset view that shows results in rows and columns for sorting and scanning.
Tags / Comments — Lightweight annotations applied to files and records for triage and collaboration (subject to permissions).
Template (dataset / record) — A configured visualization bound to a concept or dataset to standardize how results are presented.
Virtual knowledge graph (VKG) — A setup where data remains in original systems but is exposed through ontology mappings as a unified graph for Search and analysis.
Workspace — A secured container for investigations that behaves like a file system and can store evidence files, native Octostar items (Link Charts, saved searches), and workspace records.
Workspace records — Structured records visible within a workspace context, alongside unstructured evidence, to support unified investigation workflows.