With data science growing to be one of the most in-demand careers of the twenty-first century, CCR prepares students for success in the field by recommending personalized classes based on interests and prior experiences.

Berkeley, CA

UC Berkeley is very fortunate to offer over 150 different undergraduate majors and minors, which provides its students with the opportunity to experience a multitude of different fields of study. With that being said, there are almost too many options to choose from. In fact, UC Berkeley offered almost 7,000 different classes in just the fall 2020 semester alone. This makes it not only impossible to know all the classes offered but also difficult to decide. Not to mention, class selection can be especially challenging for new majors like data science, that do not yet have a set path. This often leads to an extensive amount of time spent switching between sites such as the berkeley class catalog, berkeleytime.com for grade distributions, and ratemyprofessor.com for professor ratings.

 However, a class recommendation system such as Cal Class Recommender, or CCR for short, can make the class selection process much easier. CCR provides data science related class recommendations based on classes that were enjoyed by other data science students while considering the natural order of classes. With the use of CCR, class selection is now quicker and easier than ever which is especially important when self-navigating during these times of virtual learning. CCR has future plans to expand to all majors and possibly multiple universities.

Senior student at Cal, Collin Millones, expresses his interest in utilizing CCR in his journey as a Data Science major.

CCR utilizes content-based and collaborative filtering. The recommendation engine takes in a class that a student is currently taking as its input, analyzes the contents (such as the title of the class). Then, it figures out which other users have taken similar classes. It will then rank similar students according to their similarity scores and recommend the most relevant classes to the student. For example, if the system detects that user A is the most similar to user B, then if user A has taken a class that user B has not, the class will get recommended to user B and vice-versa.

CCR was created by Chaya Bakshi and Kalina Huynh, two data science majors at the University of California, Berkeley. As they complete their journey at Cal and transition into pursuing careers in data science, they wanted to provide students with a recommendation system that will help best navigate their pathways as potential data science majors. They are dedicated to improving students’ experience at Cal, while making learning data science fun, exciting, and engaging. After all, data is the new science. 

CCR Public Repository: https://github.com/khuynh2021/CCR_data_x_f20