Carter Louchheim, Milton Vento, Rick Yanashita, Luke Young-Xu

Part 1: Needs-gathering and ideation

Our team was tasked with designing and implementing a browser extension to improve user experience online. To start our design sprint, we first had an ideation/brainstorming session to gather ideas for possible browser extensions. We jotted our ideas down on a blackboard so we could freely visualize ideas, discuss potential features/benefits, and edit them. This idea board served as a ‘working document’:

Image showing idea brainstorming on a blackboard

Brainstorming ideas on a blackboard

The extension ideas we had were

• Video capture: find videos on a web page and generate closed-captions
• Alt-text adder: add alt text to all images on a web page
• Bionic reader: make all text on a web page display in the bionic reading format

https://bionic-reading.com/br-method/

• Drink water alerter: alert the user every x hours/minutes to drink water
  • Could also work for any similar habit that requires action at a set time interval
• Automatic contrast adjuster: adjust colors on a website so the contrast ratio for normal text is 7:1 and 4.5:1 for large text (WCAG AAA standard: https://www.w3.org/TR/WCAG21/#contrast-enhanced)
• Text simplifier: simplifying complex words into words that are easier to understand for all people
• Pomodoro timer: a pomodoro-style timer extension where the web page goes dark during breaks so the user can really have a break.

https://www.forbes.com/sites/anthonykarcz/2020/03/31/try-one-of-these-5-pomodoro-timer-apps-to-help-you-stay-focused/

• Text summarizer: summarizing sentences or paragraphs on a webpage

Although we liked all our ideas, we recognized a few that were more useful and novel than the others: Video capture, Bionic reader, and Text summarizer.

We ultimately decided to pursue our Text summarizer concept. This extension will summarize highlighted text in the popup with the option for the newly summarized text to replace the original text on the web page. This way, users could either simply read the summary in the popup, or read the summarized text as a part of the original web page. To do this, we discussed that we would like to use a LLM api like the OpenAI api or Meta’s Llama api. We encountered some unexpected difficulties when implementing this, which we reflect on in the next section. To allow for the summarized text to show up on the web page instead of the original text, we discussed an implementation that finds the specific location on the page through “p” tags, and sandwiches the summary in between the start and end of the original text.