Readrr: Reading tracking and recommending app


Group project, build a recommender system to recommend similar books and keep track of user's libraries.

Published on February 02, 2020 by Claudia Chajon

recommender tracking machine learning

1 min READ

Keep track of your reading

– —

See the work


A project developed as a platform to keep track of books that are being read or have been read previously. The aim of this project is to provide a clean, uncluttered user interface that allows a user to track books, similar to the GoodReads user interface.</p>

This allowed users to rate books they had read and allow other users to view those ratings. A user may rate the books on their virtual shelf and allow for other users to view their ratings. This was a greenfield project and the data science team was tasked with setting the parameters for the project as it applies to data science.

A future goal of the project was to use NLP and predictive modeling with the reviews and book attributes (genre, topics, subject) to predict books that a user would enjoy based on their past read books.

As this was a greenfield project, the first task was to develop a dataset which could be used as a sample for a recommendation engine and use in NLP. For this reason, my group was tasked with developing a database and dataset to hold the book data.

I was responsible for the implementation of a k-nearest neighbors model on a sample database of 10K books. Recommendations were based on user ratings to find and suggest books to the user that would fit their profile.

This was part of a 2 month team project in which I worked with a multidisciplinary team made up of 6 Web Developers, 1 UX Designer and 3 fellow Data Scientists.