Guide to Conversational User Interface (CUI) Best Practices and Tools
What is Conversational UI?
Conversational User Interface (CUI) is an interface design that allows users to interact with either real humans or bots using text and voice/speech through the web, mobile and desktop applications like Amazon’s Alexa for speech, Facebook’s Messenger for the chatbot.
Conversational interfaces (CUI) are platforms that give the privilege of interacting with the computer on human language and provide the opportunity for the user to communicate with the computer in their natural language just like talking to another human, rather than in a syntax specific commands or clicking icons.
Interactions with a different conversational interfaces like a chatbot or voice assistant like Siri, Alexa, or Google Assistant, to do the shopping or to interact with your smartphone to perform some task like call to friend, search product over a website, buy a ticker or order food and many more.
The idea behind the Conversational Interface is to make communication between humans and computers easier & foster a personal connection between services and their users.
Different big IT Companies like Google, Apple, Facebook & Amazon, see conversation as the future of development. Understanding Conversational User Interfaces (CUIs) is essential to understand the future of technological development. Bots soon might get replaced with apps, enabling users to interact with their favorite digital services directly.
What is Rasa UI?
Rasa is an Open source machine learning AI framework for developers and product teams to build contextual AI assistants.
Rasa is a Conversational AI framework, not tied to a pre-built model or use case. It’s customized according to use case and requirement.
Rasa is not a rule-based framework, and no need to worry about putting data in someone else’s Cloud as in others like Dialog Flow, Microsoft LUIS or Amazon Lex.
Rasa has two main components - Rasa NLU & Rasa Core.
NLU handles unstructured inputs and converts them into a structured form that a machine can understand and acts. NLU helps in Sentiment Analysis and Conversational searches, allows a line of questioning to continue, with the context carried throughout the conversation, while Core handles dialogues and fulfillment, the goal of Rasa Core is to generate the reply message for the Chatbot.
How Conversational UI Works?
Conversational UI works just like we talk with a real human to access the different service. Interact with the system, Bot or Human using text and voice, then system responds based on queries and requirement.
To cover the basics, Conversational AI occurs in two formats —
- Chatbots (Slackbot, Facebook Messenger, Magic, and kik).
- Voice assistants or voice user interfaces (Apple — Siri, Amazon — Alexa, and Samsung — Bixby)
Overview of Chatbots
Chatbots are web or mobile interfaces that allow the user to ask questions and retrieve information from computers system. Chatbots are presently used by many organizations to converse with their users. However, with the growth of technologies like Artificial Intelligence (AI), Machine Learning (ML) and Natural Language Processing (NLP), Chatbots prominently affect UI / UX, providing complete conversational experience to the user.
Voice User Interfaces (VUIs)
VUI makes human interaction with computers possible through a voice/speech platform to initiate an automated service or process. Apple with Siri, Samsung with Bixby, Microsoft with Cortana, Google with OK Google, Amazon with Alexa, have already made VUIs familiar to the world. In future, websites powered by their VUIs or will integrate with any of the existing technologies. This will make way for more conversational website content.
Conversational UI uses -
- Natural Language Processing(NLP).
- Natural Language Understanding(NLU).
NLP to allow systems to understand, analyze and create meaning from structure to human language data and work together to handle end-to-end interactions between machines and humans in the preferred language of the human. The unstructured format of human language makes it difficult for a machine to always correctly interpret the user’s data/request, to shift towards Natural Language Understanding (NLU).
NLU handle unstructured inputs and converts them into a structured form that a machine can understand and acts. NLU helps in sentiment analysis and conversational searches which allows a line of questioning to continue, with the context carried throughout the conversation.
E.g., A user could first ask for the population of India. Then can ask “Who is the president,” the search will carry forward the context of India and provide the appropriate response.
The stack required to develop a modern and interactive conversational UI application includes -
- Speech recognition
- Conversational level
- Business logic
Benefits of Conversational UIs
- Better engagement/trust building with end-user through proper interaction.
- Better user experience/convenience & decision support as filling out a time-consuming form & adjusting questions based on the user’s answers.
- Cross-platform integration & Compatibility.
- Allows personalization.
Why Conversational UIs Matters?
- Provides realistic feel while interacting with bot & system.
- Conversation UI use words as in the natural form of communication for people, which make VUI more exciting.
- Increased user attention - targeted questions with clear Call to Action for each interaction for different users.
Best practices for Designing an attractive Conversational Interface
Building & Designing a genuinely helpful and attractive conversational User Interface is still a challenge from a UX standpoint.
GUI not works in the same way for Conversational UI works, Conversational UI Design more focuses on to design the conversation flow as naturally and efficiently as possible and focus more on words and user as compared to the visual design.
To build an attractive CUI, need to flow the best practices to design a CUI Platform, to focus on the target audience and their problems, should answer the following the question before actual implementation.
- What is the user trying to do with the system?
- Is it is solving the user/customer problem efficiently?
- Is it minimizing the user’s effort to communicate with the system?
- How beneficial is this for the target audience?
How to Adopt Conversational UI?
Guide to Clear flow
To design a Conversational Interface, Don’t try to develop system to do everything all at once, focus on only one problem/purpose. Define users problem clearly, solution, goal & expectations and the design flow should cover these points -
- Define the purpose of the system.
- Set clear expectations on what a system can do and what it cannot.
- Improve discoverability by providing hints.
- Simplify data entry.
- Offer shortcuts.
- Avoid asking open-ended and rhetorical questions.
Understanding User Control
The User should have control & freedom to use the system and need to feel in control over the system, rather than feeling controlled by system.
- Don’t leave the user waiting without providing any feedback.
- Allow the user to reset the conversation at any time during the interaction.
- Allow user interruptions & re-engage with the app after they leave.
- Confirm by asking, not stating.
- Easy error handling and error message.
- Provide help, assistance, and suggestions for when the user feels lost.
- Provide the feature to undo, redo and cancel.
The conversion between the user and system should make feel the user like talking with other humans, as people are aware that a digital system doesn’t have feelings, and prefer responses that feel warm and Humanize, rather than cold and Robotics. Be sure to design a system whose vocabulary and tone resonates target audience.
Prioritized information & Avoid jargon
Words are the significant part of Conversational Interfaces, make sentences simple, concise and clear. Use clear language and behave like conversing to real people and according to the target audience. Don’t use ambiguous language, technical terms, abbreviations, or acronyms and only show the what user wants and prioritize information according to that.
Tools, Frameworks & Platforms of Conversational UI
- RASA UI: Open source tools to build contextual AI assistants.
- Dialog flow: Build natural and rich conversational experiences.
- Wit.ai: Platform to build applications and devices that you can talk or text.
- Pandorabots: Platform for building and deploying chatbots.
- Careerbot.ai: CareerBot helps enterprises find the best talent connected to existing employees.
- Meekan: The world’s smartest AI scheduling assistant, matches everyone’s calendars in seconds.
- Manybot: A Platform to create a Telegram bot without coding.
- Amazon Lex: Conversational interfaces for applications powered by the same Deep Learning technologies as Alexa.
- IBM Watson: Go beyond Artificial Intelligence with Watson. Watson works with businesses, scientists, researchers, and governments to outthink biggest challenges.
- Microsoft Cognitive Services
- (Luis) : Language Understanding. Add conversational intelligence to your apps.
- BotMock: Botmock has everything you need to create fantastic chatbot prototypes, and it comes with the power of drag-drop editor.
- Microsoft Bot Framework: Build, connect, deploy, and manage intelligent bots to interact with your users on a website naturally, app, Cortana, Microsoft Teams, Skype, Slack, Facebook Messenger, and more.
- Motion AI: A robust platform for enterprises to build and deploy chatbots for a range of mediums, including Facebook Messenger, Slack, E-mail, Web and SMS.
- Dexter: Dexter is the platform for businesses, brands, and people to create meaningful automated conversations.
- SuperScript: Dialog system and bot engine for Conversational UI.
- Bootkit: The leading developer tool for building chatbots, apps and custom integrations for major messaging platforms.
- Reply AI: The platform to build & manage your Conversational strategy.
- Dashbot, Botsociety, Layer, Statsbot, BotMetrics, ChatterBot, Massively and many more.