EmpathiRead

• 

  Increasing the font size of the file

Inspiration

We noticed that individuals with learning and reading disabilities had struggles accessing information online due to the way it was presented. To combat this, we developed a chrome extension that flattens the barriers to access of information. This information could be anything from news to scientific articles. It can assist in bringing complex pieces of knowledge to younger people or older people alike.

What it does

This extension uses NLP to find out the user need and wants in terms of accessibility. It gathers this data and then uses custom CSS injections to change webpages into a more accessible form. It will also break apart harder terms in sentences and define them in the page for the user. We also provide the users with the ability to summarize and define terms with a very simple UI.

How we built it

We built our NLP implementation using Hugging Face. We also used LLM models from OpenAI to determine what words are difficult for our user. This API was also used for summarizing and defining terms. On the frontend we used React JS TypeScript Node.js HTML and CSS. JS and our framework was based on TypeScript using a webpack

Challenges we ran into

Some of the major challenges we ran into where that because we were developing a chrome extension there was a very limited amount of documentation and tutorials on the development and building of a chrome extension when compared to an app or website.

Accomplishments that we're proud of

• We are proud of implementing a NLP model to categorize text
• Taking years off our life with caffeine ## What we learned How to develop a chrome extension How to use TypeScript framework with a webpack ## What's next for EmpathiRead
• We will implement a audio to text and text to speech interface for our users to conversate with our extension. This will allow the application to be seamlessly and effortlessly interfaced with.
• We will also implement custom CSS injections for popular websites and partner with companies to make their websites more accessible.

Built With

• api
• css3
• html5
• huggingface
• javascript
• natural-language-processing
• openai
• react
• summarizer
• typescript