1 of 12

ProfFlow

By: Steven Peralta, Steve Huang, Froilan Zarate, Rajvir Virdi

A Chrome Extension

2 of 12

Imagine class selection time

You head to SFSU’s class schedule site to find all the classes for a specific course and time

3 of 12

Researching the class

All the information is present, but how do we know this is the section we actually want to take?

RateMyProfessor.com!

Professors can strongly influence your class experience so choosing wisely

4 of 12

RateMyProfessors is great but…

- RateMyProfessor provides great utility for research but how could we build upon this?

- Excessive information and constant tab switching creates inefficient research operations

Is there a simpler way?

5 of 12

Combine both pages!

- Utilize key information from RateMyProfessor all on the same page

- Streamlined access to essential information

6 of 12

How to implement?

- One option would be a drop down tab as an extension

7 of 12

Inspiration from “Grepper” extension

Option 2 is embedding into the page you are viewing, similar to Grepper.

8 of 12

Inspiration and Implementation from Fireworks-AI

We were inspired by Fireworks-AI’s Mixtral MoE’s ability to process vast amounts of Large Language Model data with minimal time and cost.

This inspiration pivoted our focus into faster database optimization and processing.

With the incorporation of Fireworks-AI, scaling of our work model has been streamlined.

9 of 12

What would the interface look like?

Showcases important elements from RateMyProfessors page all on a simple application
Contains the ability to read reviews with a simple mouse click
Not the correct professor? Search bar is available for manual search

The interface would show all the things a student would need to know at a glance, such as the overall rating of the professor based on how many reviews, What department they teach, what is the percentage of people that would retake the class, the estimated level of difficulty, and the spread of the ratings from one through five. If in the possibility the professor that the program detected is incorrect, as there are professors that have same or similar names, you have the option to search directly, and be redirected to the RMP page. If you would like more information on the professor, such as wanting to read the reviews that were left behind by students that have taken the class. This is the best of both worlds, giving the information needed to the student, whilst leaving the option to obtain more information if need be, whilst reducing keystrokes and tabs open, streamlining an otherwise tedious process.

10 of 12

Technicals

Key Points

Scrapes website pages for statistics
Utilizes regex command to filter out specific information
Links to RateMyProfessors.com if more information is needed
Built in Python to process information and backend services

11 of 12

Pseudocode

Prompt: “Please enter the class you want: “
input(class name)
Prompt: “Please enter the university: “
input(University name)
Use input to scrape “Rate My Professor” website (Algorithm)
Then compile data into a DataFrame (df)
Sort weighted ratings, prioritizing overall rating
store top three sorted results
Prompt: “Here are the top three professors for {class}.”
Print top three results

12 of 12

The Intended Function (Sample Output)

(prompt): “Hello, please input the class you want to take at which university: “

(User input): CHEM 101 , San Francisco State University

(Output): “Here are the top three rated professors for CHEM 101 : “

John Do Rating: 5.0

Jane Smith Rating: 4.3

Jeffrey Chan Rating: 4.2