The article was reviewed by Science X. Editorial process You can also find out more about the following: policies. Editors We have given the most importance to the attributes listed below while maintaining the credibility of the content:
fact-checked
Trusted source
Proofread
Matthew Tierney University of Toronto
![A screenshot of Aarabi’s website shows nine different versions, with each one filtered differently to represent a particular type of colour blindness. Bottom-right is weak protonomaly or reduced perception of red light. Credit: Parham Aarabi A new AI-based model can optimize online shopping sites for colourblind users](https://scx1.b-cdn.net/csz/news/800a/2024/ai-model-could-optimiz.jpg)
Parham Aarabi, a University of Toronto research assistant, has developed an artificial intelligence (AI) model to mimic how users use online shopping websites. This could help optimize retailers’ sites for those with color blindness or other conditions.
The AI-generated PRE tool lets virtual users search, browse, stop on a particular page, click discounted products, and add them to their cart.
Aarabi wanted to know how people with colorblindness would respond to the images.
Around 8%-10% of the population suffers from a form of color blindness, says Aarabi. He is an associate professor at the Edward S. Rogers Sr. Department of Electrical and Computer Engineering in the Faculty of Applied Science & Engineering. There are many ways that the eye may be misled by colors, most commonly red and green and blue and yellow.
I wanted to know how it would affect web navigation.
Aarabi conducted an experiment. Aarabi altered the website of a clothing retailer to mimic how it might appear to someone who has protanomaly or reduced perception of red light. You could think of Aarabi’s work as a lens or filter that he then altered to simulate eight different color deficiencies.
Aarabi tracked image click rates for each of the variations by initiating 1,000,000 navigation sessions using AI virtual users. Aarabi found that color-blind users are 30% more likely to click on monochrome images than color-aware users. The results of this study will be published in the paper to be delivered at the IEEE Engineering in Medicine and Biology Society’s 46th annual international conference.IEEE EMBC 2020This summer, you can enjoy a variety of activities.
Aarabi notes that the boost color gives to website designers doesn’t work for everyone.
When designing websites or presenting product, designers need to keep in mind that only 8% of people will be affected. It is important to provide more information and better text to help users navigate the buying process.
Aarabi believes that this is just one study among many others which can be benefited from PRE. It took PRE two years to develop its neural network using data collected from over 110,000 sessions of real users.
Aarabi says that to measure the accuracy of PRE, they set up an example site, predicted the actions AI virtual users would take (what percentage will add to the cart, which percentage will buy a specific product), and also tested the website with real people. PRE accurately mimics human users’ actions 90% of time.
AI virtual users can be used for research. You can conduct experiments faster, at a greater scale and in as many sessions you want. This AI model can eliminate the need to coordinate and locate thousands of colorblind volunteers.
Aarabi plans to use the PRE for testing other accessibility barriers, like dyslexia and motor impairments. He wants to provide a service that will allow companies to assess the impact of a website design on people with different conditions, before and after it is launched.
Aarabi is working to reduce the negative perceptions of AI by pursuing such goals.
He says that there is a great deal of concern, not only within the tech industry, but also amongst people in general, regarding AI replacing or taking our jobs. If we make AI more like humans, add some empathy, and let it mimic human reactions, then we might be able to dispel some concerns.
Deepa Kundur is the chairperson of the electrical and computer engineering department. She says that Professor Aarabi, who has pioneered the use of AI in the past, was a pioneer. Parham’s approach is forward-looking and valuable in leveraging AI to achieve positive outcomes.
Citation: AI model could optimize e-commerce sites for users who are color blind (2024, April 16) retrieved 16 April 2024 from https://techxplore.com/news/2024-04-ai-optimize-commerce-sites-users.html
Copyright applies to this document. Aside from fair use for private research or study, it is prohibited to reproduce any part of the document without written consent. This content is only provided as a source of information.