HookBook

Inspiration

There is a literacy crisis in America. A study in 2022 showed only 31% of 8th graders were reading at or above grade level. 83% of teachers surveyed have noticed a drop in their students' reading stamina since 2019. Reading and communication is the most important skill in the modern economy, and the American economy loses an approximate $225 Billion every year due to low literacy levels (this is $1.19T globally).

At the same time, young people are rarely reading independently due to addiction to technology, specifically platforms such as TikTok, YouTube Shorts, Instagram Reels, and more.

If we can get young people to spend as much time reading and improving their literacy skill, as they do mindlessly scrolling through online content, they will rapidly improve their skill and subsequently their future success and quality of life.

What it does

HookBook makes reading addictive by making reading more like TikTok. In HookBook, users swipe and tap through an infinite feed of engaging 'hooks' about short stories. We have 1,000 real short stories from an open data repository that we have rewritten using AI to be accessible to young people, while keeping the plot and characters as dynamic and relatable as their authors have made them.

After readers find a hook that catches their interest, they tap to read more, they can ask questions about the story as it unfolds, and when they finish the story, they answer comprehension and discussion questions to show what they just learned.

Then, our recommendation algorithm learns the student's interest based on which stories they pick, and tailors the next story shown to those exact interests.

How we built it

This was built using an open source dataset of short stories, a starter Django template we found–from which we customized to suit our needs, Figma for design work, Cursor for assisted development, and a plethora of other minor frameworks and packages.

Challenges we ran into

Some challenges we ran into include finding the right short story data set, and then fine tuning our ML re-writes of them, creating a cosine similarity recommender system with limited resources, and deploying a full stack app on a tech stack most of us weren’t familiar with. Our idea is big and so are our responsibilities.

Accomplishments that we're proud of

Some accomplishments that we are proud of is being able to support a working system that carries out the intended task. Trying out the product for ourselves was not only fun but amazing as well. HookBook serves its purpose as encouraging reading amongst teens who are constantly living within the digital world.

What we learned

We learned a lot of things at HackHarvard. One primary thing we learned is that it takes a lot of trial and error with figuring out how to create a product that solves a complex problem with an efficient solution. Although we communicated and cooperated thoroughly throughout, we also learned the value of self management and specialization. We also learned a lot of technical skills ranging from going from no knowledge of Django to writing a full stack application.

What's next for HookBook

Future plans would be to contact schools and parents to sample current features. Within a year, we expect to see significant growth and feedback without our web app to have a sense of most demanded features. One feature we plan on implementing soon would be to have small and simple AI generated images that match each story and have them appear along the text as the child reads. Another feature we plan to include is augmenting the recommender system with user input (such as likes for good hooks). And last among the many other features we plan to implement is to tag stories with genres and allow users to explore the stories according to their desired genre.