Research Paper Assistant (Beginner) (Startup) (Overall)

Inspiration

Undergraduate students have struggled to read research papers. All three members of our team have all spent hours reading research papers that we were not able to understand.

What it does

It provides the user with a webpage to view their research paper pdf with ai generated definitions for words which may be hard to understand.

How we built it

We built it by using 3 python files which do different functions each. The first one retrieves the pdf data and adds it into the html website. The second file prompts an AI model api to give us our request that we ask the chatbot. We then send this message to the third file which adds the popup definitions to the html webpage. We use a storage txt file to store the pdf data and properly edit it.

Challenges we ran into

One big challenge we ran into was getting the ai model to give the correct responses we wanted. We solved this by doing trial and error and giving small sentence inputs to the chatbot to get the optimal response. Even when some words were single or 2 letter words which would mess up the html file, I edited one of my loops in main.py to avoid words under 3 letters.

Accomplishments that we're proud of

One thing that we are proud of about our project is how we avoided retrieving unwanted text from tables effectively. This will allow the reader to view the research paper more clearly.

What we learned

We learned that parsing through text may seem like a simple process but it was the hardest part of the process. Accounting for many different types of pdf files is a hard process and is the next step in our project.

What's next for Research Paper Assistant

What is next for Research Paper Assistant is to incorporate more types of pdf files which remove the footers, headers and references which makes the user experience better.