Inspiration
Libraries are one of the last few ways to access entertainment, information, and technology for free, yet are at risk of being defunded these days. With this project, I want to show that many people benefit from the services offered by SPL.
What it does
I made a list of the top 10 most frequently checked out books, a graph of how book checkout frequency has changed over time, a statistical significance test for the drop in checkouts since COVID, and two (inaccurate) machine learning models to predict checkouts for a book.
How we built it
I filtered the data to just look at books and ebooks and just checkouts from January 2016 to February 2022. To make the top-10 list, I grouped the data by title and sorted by checkouts in descending order. For the graph, I used seaborn to graph total number of checkouts per each month. The statistical significance test was from the scipy.stats library, and to make the machine learning models I used month and year of checkout as features and number of checkouts as labels.
Challenges we ran into
The size of the data was a big challenge. I was not able to use some of the variables I wanted to use, like subjects and publication year, because pre-processing these would be very inefficient. For example, when I tried to do one-hot encoding on the publication year I got an error because the dataframe that was returned was over 60 GB. This limitation in the variables I could use made my machine learning models very disappointing.
Accomplishments that we're proud of
I am proud of my graph of checkouts over time, and of having figured out how to do an independent t-test with scipy.
What we learned
Library usage is on the rise in recent years. Also, the most popular books of Seattle readers tend to be modern books that discuss social issues.
What's next for What Does Seattle Read?
I could try redoing the project with a smaller section of the data so that pre-processing is more manageable. Also, I'm curious about whether checkouts have increased in proportion to the Seattle population, and about how the proportion of Seattle residents with an SPL card has changed.
Log in or sign up for Devpost to join the conversation.