Skip to main content

Glossary

Main concepts​

Knowledge Graph / Ontology / Mapping (Octostar Fusion Center)​

Octostar can operate on petabytes of data, based on the principle that it can connect to existing sources of data and map them to an existing data model (ontology).

true.png)

https://app.diagrams.net/#G1KC4DQ_9X0UPY4TiUiVh4u9rcKSpy4rqX

Ontology

A class/relationship based structure which captures the domain we're operating in. Very similar to Object Oriented class hierarchy (or RDFs Schemas).

The Octostar Ontology, is a specific ontology developed by Octostar which contains hundreds of concepts and relations, ready to use. This might be the one used in your Octostar deployment.

Concept

Term inherited from Timbr, stands for "Class" in the typical semantic web world (RDFS/OWL) but also in classic object oriented

e.g. "SoccerPlayer"

Attribute

It is a field of a class, e.g. DateOfBirth . In timber they can be simple or "multi value"

Relationships (data model level)

Schema level definition of the possible relations between classes (they can be different or the same class e.g. (Player) "playsIn" (Team)

Mapping

Something that maps a physical table on a raw datasource to an Ontology Concept e.g. a table "customers" to a concept of "users" subclass of "persons"

Knowledge Graph

A combination of an "Ontology" and set of "Mappings". Things make sense and have relationships among each other only within a context of a specific Knowledge Graph.

Schema?

The reasoning engine used. TBD

Semantically Bound

Within Octostar, a script, a template, a dashboard, a webservice driver can be said to be semantically bound when they specify one or more concepts they are meant to handle.

For example one can create a specific Record and Set Templates and then specify it applies to "People". In this case thanks to this setting it will be automatically used to visualize people (or shown as one of the alternatives)

Entity/Record level terms​

Record (Entity)

A specific instance of a Concept. e.g. an Record of "SoccerPlayer" is "Neymar" or a specific car with it's own serial number (note not a record.. ). Note that while "Entity" might be the more correct "Semantic" term and might sometime be used interchangedly - within Octostar we prefer the use of the term "Record"

Attribute Value

The specific values of the attributes e.g. the Attribute "color" for a specific car is "red"

Relations

A specific instance of a relation between two entities.

Application level term​

A Workspace

A partitioning for items of certain types of data so that it is only visible to those that have access to that investigation.

WSFS - the Workspace File System. Like a regular file system contains items and folders.

A Workspace item

An item that is contained in a workspace. These can show up in the WSFS (WSFS Item) or not.

WSFS Items​

A Graph

A document type where nodes and edges are place in a specific x/y position. It is serialized and lives within an investigation space.

A Report

A Search

A Folder

A pointer to a record

A Binary

Non WSFS Items​

The desktop​

The desktop refers to the CURRENT status of a user session in term of which workspaces are currently open, and therefore which APPs are currently loaded, but as well which specific windows are currently open.

Data Viewers / Dashboards etc.​

Summary overview:

To have "information" about a set of records (e.g. 50k people) the user double clicks on the sidebar concept tree.

This summons a "dataset viewer" tab.

This has:

  • a filter bar, which shows filters currently applied to the concept. e.g. all "blonde" people. filters are emitted by Dataset Views (see below).
  • A Dataset View switcher tab. This lists all the dataset views (see below for definition) which are available. Most of these are dashboards which have been "associated" with a concept in the datamodel. E.g. there could be 3 dashboards which have been associated to the concept of "person" then these would all show in the Dataset view switcher bar.
  • Lastly under the switcher bar, there is the content of the specific Dataset View that is currently selected - eg. the visualizations of the specific dashboard which is acting as Dataset view at this time.

Formal definitions:

Filtered Concept (AKA "Search")

A Filtered concept is a concept plus a set of filters.

Example of "Filtered Concept":

  • People AND "age<43"
  • Cars AND (Licenceplate=A432HF OR Licenceplate=B458JGF OR Licenceplate=B458JGF. .... )
  • Events AND timestamp<15 minutes

Note: users will tend to call these "Searches" as a Filtered concepts can be cerated from the "Search" functionality (or from starting from a concept and drilling down).

Dataset Viewer​

Takes a Filtered content and provides a UI with:

  1. A bar with filters - where the current dataset specification is clearly visible - e.g. can be a filter bar and an ICON + NUMBER e.g. 54645 PEOPLE hair=blonde

  2. A bar to switch Dataset Views

Dataset View​

An individual view that can visualize a dataset (concept + filter)

Typical example is a dashboard that is semantically bound so it can show "people".

But can also be:

  • A Template (which is marked as supporting sets)
  • An App (which is marked as being usable as a set view)

Dashboard (simple)​

A simple superset dashboard. Saved in the WSFS

Dashboard (Concept bound)​

A dashboard which has been bound to a concept. This allows the dashboard to act as a Dataset View

Individual Record Viewer​

This is what one gets when clicking and opening an record(E.g. "open in record viewer" from a graph or a list of records.

Typically it shows the record via one of the available views e.g. the main record viewer (Where one can also edit) or any of the record templates Record and Set Templates or semantically bound apps