37th Annual CSUN Assistive Technology Conference Has Concluded

Accessible Conversational UI’s

Date & Time

Wednesday, March 16, 2022 - 2:20 PM PST

Location

Orange County

Description

Conversational UI’s help answer questions from customers. Learn to define solid accessibility guidelines that can help ensure that the inefficiencies of the CUI don’t limit the users it caters to, especially in this new normal of “virtual assistance”.

What are some examples of a CUI?

Chatbots are a common form of CUI with a substantial history of development. Common usage are Customer service, entertainment and therapy. Chatbots are often interacted with through text chat, but spoken dialogue is also possible.

Voice Assistants sometimes also called as VUI (Voice user Interfaces) or VAPA (Voice Activated personal assistants) are virtual assistants that operate through dedicated hardware devices such as Amazon Alexa, google home

Virtual assistants such as Apple’s siri, Google’s assistant and Microsoft cortana's are very popular.

IVR are a further form of mainstream of CUI, which are commonly used in telecommunications systems where organizations deal with large call volumes. These tend to be relatively limited, commonly using pre-recorded audio and specified response options.

Next, we will dive into the Key features of a CUI

Potential for multiple communication modalities -
Dialogue is fundamental to CUIs, but this should not restrict the CUI to use just one communication mode for example, text or speech. There are propositions that say CUI’s should include non-verbal communication as well because conversation can include elements such as body language.

Potential for use through multiple channels -
A channel refers to third party applications. For example there are existing popular messaging platforms like Messenger, Slack where people spend a considerable amount of time. These are platforms where chatbots are fundamental to interacting with audiences. One thing to keep in mind is for making sure these are accessible, it is very essential to understand how accessible the original application itself is.

Dialogue based interactions -
CUIs are primarily based around exchange of utterances in a dialogue. Interaction with a CUI follows a similar structure to the command line. The choices or actions available may be restricted and less apparent to the user of a CUI v/s a more complex ‘mixed initiative dialogue’ by both the system and the user.

Combining logic and machine learning -
Many modern CUIs combine rule-based logic and statistical machine learning algorithms. As with other AI systems, bias introduced by the data used to train the AI, or in the algorithms developed may create inequities between users. These can be problematic for users with a specific disability affecting a minority of users, who may not be represented in the training data or in the design process.

This Presentation Link is provided by the Presenter(s) and not hosted by the Center on Disabilities at CSUN. The Center on Disabilities has confirmed, as of April 1, 2022, content linked is relevant to the presentation, but has not been reviewed for accessibility nor will the Center on Disabilities attempt to remediate any accessibility issues in the linked content. Please contact the Presenter(s) with any accessibility concerns.

Audience

Information & Communications Technology
Research & Development

Audience Level

Intermediate

Session Summary (Abstract)

In this world of “live help, a click away”, it is critical for organizations to ensure that their CUI’s are accessible. This is how they would truly “solve for ALL their customers”. Intuit is working to ensure that our live help products are accessible. We will share our learnings with you.

Session Type