Salisbury University
https://www.salisbury.edu/
The home of The Rommel Center Research Concentrations Study. This project harnesses various AI tools to analyze textual data and create an application that will visualize each category of research at Salisbury University. Each individual category will have in depth details related to funding, Influential Faculty, etc...
- AI analysis of different documents to find keywords, topics, and recurring themes which is fed into a database
- User Interface with categorical data visualization and the ability to drill down each category into detailed chunks of data.
- Comprehensive Map of all research being done at Salisbury University with a user friendly interface and ease of use.
- There should also be a search feature to find exact results
- At the very least, pages associated with each concentration should have enough data to conclude if a topic is a strong or weak area of research at SU
The main issue with interpreting various sets of text is the context window. Open AI's GPT 3.5 Turbo allows for about 4,096 Tokens in it's context window which translates to about 1,000 to 1,500 words. Solution: Automate the process of parsing blocks of text that fit into the Context Window of GPT. Then ask for the keywords and topics throughout the text and build a program around this that will keep all the information together.
- Tailwindcss for styling.
- React to make modularity possible and more simple to use with javascript.
- JavaScript
- Python
During this semester, We will be completing a spreadsheet that allows us to first gather the needed data and organize it for businesses at minimum.
The team has decided to use a pretrained bert model and various textual analysis tools to derive the output topics and sub topics
- set up git
- python basics
- Design Document
- Tensorflow
- powerpoint
- Basic ML understanding
- BERT
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent