ShelfSprout

AI-Powered Book Club and Local Bookstore Connector

Qadira Razman, Ranju Krishna

Link to Project Report: https://docs.google.com/document/d/1p0PMcRHc1eUaX7L4ISCz7cMXmoBB-CceJOnsMxqCebM/edit?usp=sharing

Project Brief:

Target Learning Audience:

Book club organizers, avid readers, independent bookstore owners, and small business owners. The audience would also encapsulate women and non-binary folks since this community would function as a safe space for them.

Identified Learning Need:

Book clubs have the ability to play a vital role in fostering community engagement and literary appreciation. However, to create and execute a successful book club is easier said than done. Book clubs actually have the potential to be a ‘high-leverage pedagogical strategy’ (Polleck, 2024). They offer alternative spaces that can be simultaneously safe and intellectually challenging. Many book club organizers struggle with logistical challenges, such as finding accessible places to meet, sourcing books affordably, and discovering local businesses and events that align with a range of reading interests. Independent bookstores, on the other side of the coin, are constantly searching for ways to attract and engage readers but they might lack the digital tools to connect with book clubs in the area. Book clubs have been known to allow individuals to not only strengthen their emotional connections to the stories at hand but also with others (Stone & Maimone, 2024). This speaks to the ability of a book club to enhance a stronger and more meaningful sense of community engagement by providing these curated literary experiences. In bridging this gap between book clubs and bookstores, we have the opportunity to strengthen literary communities while supporting small businesses.

Rationale for AI Assistance:

AI would aid in creating a curated, personalized recommendation system for book club groups to find their next book and vendor, as well as finding solutions for locations and logistics for meetings. These tasks would take longer or contain inaccuracies if done by humans, so introducing AI as the solution to this problem would create an efficient system for both book clubs and bookstore vendors. In terms of the book club users, AI could also assist in creating discussion questions for groups, suggesting activities that relate to the chosen novel that the club can partake in, and many more instances. AI could become a helpful facilitator in creative ideation for book club users in this situation. Additionally, AI would assist in somewhat complicated tasks for indie bookstores such as inventory management that reflects changes made through bookclubs selecting them as their store, and track overall revenue and suggest any areas of improvement. Prior research has been conducted on the usage of AI to empower small businesses, with many studies reporting that AI can optimize inventory management “with up to 70% cost reduction” (Demirci, 2024). Employing AI for this product would solve the problem of inaccurate recommendations from existing products and help ShelfSprout become a powerful tool for small businesses to create a steady stream of revenue, and foster a tight-knit community between bookclubs and bookstores.

References:

Polleck, J. (2024). Engaging Minds Through Literature. literacyworldwide.org Stone, S., & Maimone, J. M. (2024). Crafting a Pandemic Book Club. The Library Quarterly., 94(2), 202–218. https://doi.org/10.1086/729232 Demirci, S. (2024, March 27). Empowering Small Businesses: The Impact of AI on Leveling the Playing Field - Orion Policy Institute. Orion Policy Institute. https://orionpolicy.org/empowering-small-businesses-the-impact-of-ai-on-leveling-the-playing-field/

Ethical and Societal Implications

While our project works with AI to generate a solution for book clubs and bookstore vendors, we wanted to make sure it is used in a responsible way that uplifts everyone.

In terms of data protection, we plan on being transparent with how their information is being used and stored, especially for bookstore vendors that are inputting sensitive information such as their current revenue and financial budget. The data that is being collected will be the bare minimum to create recommendations, whether that’s for book clubs, readers, or businesses. We will also offer the option for users to delete their data at any time, giving them autonomy over their data and including them in this process as much as possible. With the aspect of data protection, our goal is to take an ethical approach to working with the user rather than for them.

For societal implications, the main thing we wanted to focus on is avoiding bias and discrimination in the book club recommendations of our project. Since the user would be sorted based on their responses to personality questions and reading interests, there is a chance that an AI-trained model would group responses based on stereotypes. For example, there’s a chance the model could develop the idea that South Asian authors will be recommended to only those who “seem” to be South Asian. This is a harmful stereotype and would also close off readers from learning more about the issues and culture of this minority. To mitigate this, we wanted to make sure the data being used comes from a variety of sources, whether on the internet or target user data from surveys. We would also be supervising the data throughout the process, to ensure that there are added levels of protection for our project from biases. To involve users in this process, we plan on including a user feedback feature that allows users to submit any issues or errors they see in the final product, which will be checked regularly.

Problem Statement:

Book clubs have the ability to serve as powerful spaces for community engagement and intellectual discourse, yet organizers often face logistical challenges such as finding accessible meeting locations and fostering meaningful discussions. Meanwhile, independent bookstores often seek to attract and engage readers but might lack the digital tools to connect with book clubs in the area. This disconnect limits opportunities for collaboration and local business support. ShelfSprout, an AI-powered platform, aims to bridge this gap by providing personalized book recommendations, facilitating book club logistics, and strengthening connections between book clubs and local bookstores. ShelfSprout aims to empower both book clubs and small businesses, nurturing sustainable literary ecosystems.

Key Findings From Our Need-Finding Report:

Frustrations in Community Building

A big struggle for many book lovers is the difficulty of finding and joining book clubs that align perfectly with their reading and personal preferences. We were able to deduce that the real difficulty lies in forming meaningful, real-life connections with the reading community. Book clubs often lack in terms of providing streamlined information on their meeting-styles and offerings, making it hard for potential members to decide what the right fit for them is.

Need for Curated, Personalized Book Selections

Users expressed a strong desire for personalized book recommendation based on reading history and general preferences. Current discovery methods may require extensive searching and may not always result in satisfying results.

Accessibility and Awareness of Local, Independent Bookstores

Users shared their desires to support independent bookstores but faced challenges in locating them (this might be due to outdated websites or a lack of information online).

Must-Haves:

  • AI-powered book and book club recommendations via a matching system based on user preferences.
  • Enhanced bookstore discovery and connection tools.
  • Community-building features that aid in facilitating in-person and online discussions.

Click here to see our full need-finding report!

Click here to see our survey!

For our current prototype of the project, click here: https://e412493a-92d2-41e9-85ce-d6c421744795-00-2r3ip5g6luofc.picard.replit.dev/