Bixby Developer Center



This glossary covers the main technical concepts discussed in the Design and Developers' Guide.

Video Tutorial: Cross Platform Voice Development Terminology with Bixby, Alexa and Google Assistant

The following video gives an overview on the various terms used when developing for different voice assistants.



An abstract definition of some operation that might be performed. It can be a search (such as FindRestaurants) or perform an operation (such as BookHotel), and contains action inputs and one action output. For more information, see Modeling Actions.

action input

A description of an input of an action; in particular, it includes a name, a concept, and a cardinality. For more information, see the input reference page.

action output

A description of the output of an action; in particular, it includes a concept and a cardinality. For more information, see the output reference page.

action type

A description of how an action can be used, including whether it has side effects and whether its effects on the world are atomic or continuing. For more information, see the type reference page.


The output of a strategy. For more information, see Selection Learning.

aligned text

A format for NL training that includes the utterance as well as embedded annotations (or labels). It is used internally by the platform to represent training provided by developers. For more information, see Training and Vocabulary.


A user can provide authorization for a function to access an external provider via a prompt. For more information, see Authorizing Access With OAuth.



When extending a structure concept, a particular property can be bound to a specific value (for example,Restaurant can extend a Business and bind the business category property to be "Restaurant"). For more information, see Relationships Between Concepts.

binding path

A dot-delimited set of property names indicating a property or any depth of sub-property of a structure concept (for example, on a geo.LevelOneDivision or review.rating.score on a restaurant.Restaurant). For more information, see Structure Concepts

Bixby service

Once a capsule is deployed to the Marketplace, it becomes a Bixby service. End-users add and use Bixby services within Bixby.



A unit of contribution in the Marketplace, consisting of a group of related functionality. It also delineates permissions among developers. For more information, see the Overview. Once a capsule has been reviewed, accepted, and deployed to the Marketplace, it becomes a Bixby service.

capsule lock

A period of time when the user is locked into the current capsule before it times out. This ensures that subsequent utterances are always sent to the current capsule for a short amount of time. For more information, see capsule lock.


In many cases in modeling (such as describing action inputs), it is necessary to describe how many values can appear. Cardinality is described in terms of "min cardinality" (optional or required) and "max cardinality" (one or many). For more information, see Cardinality.


"Compile" is used in two senses in Bixby: compiling your capsule's code as a whole, and compiling your capsule's Natural Language model. Compiling your capsule's code happens automatically, but you must explicitly compile your NL model before Bixby can use it to understand utterances. This functionality is available in the Device Simulator, the Story Editor, and the Training Editor.


Modifications or additions by developers to code or data that affect operation of the system, including models, functions, dialog, layout, and NL vocabulary and training.

confirmation prompt

A prompt that asks the user to confirm something, such as making a purchase. For more information, see Confirmation Prompts.

conversational interface

A user interface in which the user exchanges queries and responses with the platform in a dialog, until an answer is reached. For more information, see Designing Conversations.


An abstract definition of a class of possible values. Anything that a user can refer to or that a developer wishes to model can be a concept. Concepts include primitive concepts and structure concepts. For more information, see Modeling Concepts.

concept value

A data value that is an instance of a concept.

core models

See library models.



A point in time, consisting of a date and a time. Dates and times can also be used in isolation. For more information, see DateTime (viv.time).


A person or organization that can make modifications to capsules.


A description of the responses the platform can make in response to a user, or while fulfilling a request. For more information, see Refining Dialog.

dialog event

A dialog response triggered before, during, or after execution. Examples include progress notification, action confirmation, prompts for more information, error messages, and result messages. For more information, see Dialog Events.

dialog fragment

A composable template for dialog, relevant to a particular mode. The modes are Action, Concept, Input, Sort, and Value. For more information, see Dialog Fragments.



An operation that changes the plan in response to conditions encountered during execution. Effects include replacing or dropping an input, prompting a user for new input, or halting plan execution. For more information, see Effects.

enum concept

A structure concept that has a fixed set of values. (For example, the viv.measurement.LengthUnit concept might have values Inch, Foot, Millimeter, etc.) For more information, see the enum reference page.

equivalence definition

A description of the conditions for when two concept values might or might not be the same. Equivalence can return "true," "false," or "uncertain". For more information, see Merging and Equivalence.

Expression Language (EL)

A simple language for expressions used in a few places such as dialogs, strategies and layouts. For more information, see Using Expression Language.


The process of running a plan to produce a result, and the associated presentation to the user. For more information, see Planner Overview.

execution graph

The resulting directed graph that outlines the plan Bixby used during execution. For more information, see The Execution Graph.


A relationship between two concepts, where a child concept can extend a parent concept by adding additional properties. Represents a single-directional is-a relationship (for example, a restaurant is a business). For more information, see Extension.



An addition to a conversation from Bixby after giving an initial result. For more information, see Guiding Conversations.


An implementation of an action (typically defined in JavaScript). For more information, see Implementing JavaScript Actions.



See goal signal.

Goal signal

A signal in a plan intent indicating the result type of the plan. For more information, see the goal reference page.


hands-free mode

User-mode of Bixby that is activated when users say the wake phrase "Hi Bixby". Normally, this implies that users cannot see and/or touch the device to interact with it. If Bixby is presenting options in a selection or reading out results from a list, then the user is using hands-free list navigation while in hands-free mode. Additionally, the $handsFree expression variable returns true in this mode. Bixby-enabled watch devices can only be put in hands-free mode when they are paired with Bluetooth headphones.

hands-on mode

User-mode of Bixby that is activated when users press and hold the Bixby button on the device or taps the Bixby button on the device's screen when Bixby is already opened. This implies that the user can see and/or touch the screen, as well as use the Bixby button to answer. Additionally, the $handsFree expression variable returns false in this mode.


instantiation strategy

A strategy that suggests default values for a concept for an instantiation prompt. For more information, see Default Values.


The process of converting an utterance to an intent.


A formal expression of what a user's expressed intent. It is made of signals. For more information, see the intent reference page.



A description of the way content or forms are presented to the user within a larger views file. For more information, see layouts.

library capsules

Capsules that you can import into your capsule that provide essential models and actions, such as geography and time. For more information, see Library Capsules.

library models

Certain essential capsules that are provided by the platform through library capsules, including support for geographic models and data, date and time, payments, and credentials. Also known as code models.


A specific location defined by an ISO 3166 code, made up of a two letter language code and an option two letter country code. For example, en-US is the US English locale, en-GB is the Great Britain English locale, en is the English locale that covers all english-speaking countries, and ko-KR is the South Korean Korean locale. The base subfolder in your resources directory covers all targeted locales. For more information, see Organizing Your Capsule Resources.


match pattern

A pattern that matches against a subset of the plan graph. It is expressed as a labeled sub-graph of the plan graph, written as a tree where the root is the current point of execution.


The Marketplace is the multi-device, multi-locale location, where all your capsule submissions can be published to all the Bixby targets defined in your capsule.bxb file. For more information, see Working With the Marketplace.


In a test set, a mismatch indicates that the received goal, the goal predicted by Bixby's planner for an Aligned NL utterance, did not match the expected goal, the goal specified in the Aligned NL.


A set of abstract definitions of concepts and actions. For more information, see Introduction to Modeling.


The definition of concepts and actions to enable the platform to solve a use case. Good modeling takes into account the planning and execution process.


The response from Bixby within a conversation that is made up of dialog, views, and follow-ups. There are three types of Moments: input moment, confirmation moment, and result moment. For more information, see Moments.



The action models and concept models within the larger execution graph created by Bixby to fulfill a goal. For more information, see Planner Overview.

NL training

Examples contributed by developers to that teach the platform's machine learning system how to map utterances to intents. Internally, NL training is represented as aligned text, though more commonly it is edited through the training editor.

Natural Language (NL)

Anything a human might type or say when interacting with Bixby. It need not be grammatical, and can be of any language.

node ID

A unique identifier for a node. The node ID of a concept or action is typically of the form air.Flight or restaurant.FindRestaurants (where, in the first example, air is a capsule name, and Flight is a name for the concept, and similarly for the action). Other nodes might have internal or generated IDs that reflect their function in a graph.


outer goal

A goal signal that labels the entire utterance, and is not attached to a particular word. For more information, see the @context reference page.


When extending a concept, it is possible to override a property with modifications. For example a Restaurant concept can override a Business concept's name property to be a RestaurantName. For more information, see the override reference page.


An entity that has ownership of concept values, including both capsules and users.



The ability to access to a specific piece of sensitive data, such as a credit card number or password. For more information, see Grant Capsule Permissions.


The core software of Bixby, excluding contributions from developers, including systems that support interpretation, planning, and execution, as well as library models.


A set of steps (technically, a graph) that might be executed to yield a goal. A plan can include a goal, route, reference, or subplan. For more information, see Program Generation.


The component of the platform that produces a plan based on an intent. For more information, see Planner Overview.


The process of producing plans from intents.

plan graph

The graph that describes a plan, consisting of nodes corresponding to nodes and edges indicating relationships, such as inputs, outputs, projections, and other operations.

primitive concept

One of the primitive data types, such as numbers or string names. For more information, see Primitive Concepts.


An interruption in execution to confirm with the user, or to elicit a value or selection. For more information, see Prompting Users.


A field within a structure concept. Represents a "has-a" relationship between the structure concept and the property's concept. For more information, see Properties.


An entity or organization that provides data or services to the platform. For example one provider might be a weather data feed; another might be a restaurant booking API.


recovery goal

A type of intermediary goal for which a prompt is prompting. When Bixby fails to reach the original goal and has to prompt the user, that response is used to reach the recovery goal, which in turn helps Bixby complete the original goal. For more information, see Prompting Users.


An inquiry by the user to Bixby. For more information, see Requests.


A concept value that is the output of a function.


A signal in a plan intent indicating a concept or action that is a way-point in the plan. For more information, see Routes.


selection strategy

A way to offer advice about which values of a selection prompt are good and which ones are not. For more information, see Selection Strategies.

selection prompt

A prompt that asks the user to make a selection. For more information, see Selection Prompts.


See web service.

side effects

When an action changes things in the world, such as modifying things remote service, or performing an action in the "real world," we say it has side effects. For more information, see Transactional Workflows.


A part of an intent representing a node and the role it is expected to play in the desired plan. For more information, see the intent reference page.


A suggested approach for handling execution flow in a particular situation, such as an instantiation strategy or selection strategy. For more information, see Personalization and Learning.

structure concept

A concept that, like a record, holds multiple properties, each itself corresponding to another concept. For more information, see Structure Concepts.

subplan signal

A signal that indicates a recursive subplan. For more information, see the subplan reference page.


Any string value, when used to define a primitive enum concept. For more information, see the enum reference page.



A device and locale collection of capsules. For example, bixby-mobile-ko-KR is the target for all capsules for mobile devices in Korean. For more information, see Set Device and Locale Targets.

test set

A file consisting of Aligned NL lines used to evaluate the accuracy of a capsule's training examples. For more information, see Training Evaluation.


Referring to the sum total of all NL contributions for a capsule. For example, you could say “The training for the Uber capsule is in great shape”. But you shouldn't say “The trainings for the Uber capsule are in great shape”. See also NL Training. For more information, see Introduction for Training for Natural Language.

training entry, training example

An individual unit of training for a capsule. A set of those units would then be referred to as training examples (not "trainings"). See Training for Natural Language.



An end-user of the entire system.


A natural language request made actually or potentially made by user.



See concept value.

value prompt

A prompt that asks users to provide a value. For more information, see Prompting Users.

value signal

A signal in a plan intent indicating a concept value. For more information, see the value reference page


Renders and displays the content of a Moment. For more information, see Building Bixby Views (UI).


Words or phrases contributed to a particular node, to allow NL intent interpretation to recognize when that node is referred to. For more information, see Extending Training with Vocabulary.


web service

A service used by function implementations, such as a REST or SOAP web service. For more information, see Calling Web Services. This is different than a Bixby service.