Dialogue management a How the system manages the conversation between the healthbot and the user.	Finite-state: The user is taken through a predetermined flowchart of steps and states. If the user veers from the path, the system is unable to respond.
	Frame-based: The user is asked questions and the system fills slots in a template to perform a task. If information regarding multiple slots on the template is given, the system interprets both. Thus, the dialog flow is not predetermined but rather depends on the user’s input and the additional information the system needs.
	Agent-based: This enables complex communication between the system, the user, and the application. This uses advanced statistical and AI methods to manage the conversation.
	Undetermined: The app’s dialogue management was not able to be determined.
Dialogue interaction method b Which method the healthbot employs to interact with the user in the conversation.	Fixed input: The system is responsible for managing the state of dialogue during interactions between the agent and the user. It characteristically uses "button-push" interfaces with finite responses. It typically does not involve NLP techniques.
	Basic parser: The system parses and computes the input to decide on a final reply to the user. It can only respond to basic questions.
	Semantic parser: The system uses a flexible plan so dialogue-based interaction can be dynamically calculated based on information the system gathers about the user. It can answer a wider range of questions as it is not restricted to keywords. It is also able to glean themes from past user questions and dynamically respond accordingly.
	AI generation: The system generates replies to users through machine learning algorithms or statistical approaches. It can respond to complex questions of 2–3 sentences.
	Undetermined: The app’s dialogue interaction method was not able to be determined.
Dialogue initiative a Who initiates the conversation between the healthbot and the user.	User: The user leads the conversation.
	System: The system leads the conversation.
	Mixed: Both the user and the system can lead the conversation.
	Undetermined: The app’s dialogue initiative was not able to be determined.
Input modality a How the user can communicate with the healthbot.	Spoken: The user must use spoken language to interact with the system.
	Written: The user uses written language to interact with the system.
	Visual: The user uses visual cues (e.g., graphics) to interact with the system.
	Undetermined: The app’s input modality was not able to be determined.
Output modality a How the healthbot system can communicate with the user.	Spoken: The system uses spoken language to interact with the user.
	Written: The system uses written language to interact with the user.
	Visual: The system uses visual cues (e.g., graphics) to interact with the user.
	Undetermined: The app’s output modality was not able to be determined.
Task-oriented a If the healthbot system is intended for a specific task, or is intended purely for conversation.	Yes: The system is designed for a particular task and thus engages in short conversations to determine the necessary information to accomplish this set goal.
	No: The system is not set up to fulfill a short-term goal or task.
	Undetermined: The app’s task orientation was not able to be determined.

^a(adapted from Laranjo et al.).

^b(adapted from Montenegro et al.).