Board 97: Work-in-Progress: TextCraft: Automated Resource Recommendation for Custom Textbook Creation

Xinyuan Fan; Hamid S Timorabadi; Salma Emara

Download Paper | Permalink

Conference: 2024 ASEE Annual Conference & Exposition
Location: Portland, Oregon
Publication Date: June 23, 2024
Start Date: June 23, 2024
End Date: July 12, 2024
Conference Session: Electrical and Computer Engineering Division (ECE) Poster Session
Tagged Division: Electrical and Computer Engineering Division (ECE)
Permanent URL: https://peer.asee.org/48398

Request a correction

Paper Authors

biography

Xinyuan Fan University of Toronto

visit author page

Xinyuan (Elva) Fan is currently pursuing a Master's degree in Electrical & Computer Engineering at the University of Toronto, following her Honours Bachelor in Computer Science from the University of Waterloo. At the University of Toronto, she worked on a research project focusing on web crawler-driven automated textbook creation. She can be reached at elva.fan@mail.utoronto.ca or elvafan625@gmail.com.

visit author page

biography

Hamid S Timorabadi P.Eng. University of Toronto

visit author page

Hamid Timorabadi received his B.Sc, M.A.Sc, and Ph.D. degrees in Electrical Engineering from the University of Toronto. He has worked as a project, design, and test engineer as well as a consultant to industry. His research interests include the applicati

visit author page

biography

Salma Emara University of Toronto

visit author page

Salma Emara is an Assistant Professor, Teaching Stream in the Department of Electrical and Computer Engineering at the University of Toronto. She received her B.Sc. in Electronics and Communications Engineering from the American University in Cairo in 2018, and her Ph.D. in Computer Engineering from the University of Toronto in 2022. Her Ph.D. research focuses on improving reinforcement learning algorithms to solve problems in computer networking algorithms. Currently, she is interested in building software-tools for programming education and pedagogical practices that build testing and debugging skills for beginner programmers.

visit author page

Download Paper | Permalink

Abstract

In the educational landscape, textbooks have long been fundamental pillars for transmitting knowledge. Traditional textbooks often fall short in addressing the unique needs of various courses and the ever-evolving nature of academic disciplines. They lack the flexibility instructors need to tailor materials for specific course requirements or a particular semester. This lack of flexibility makes instructors face the dual challenge of ensuring their course content remains both up-to-date and aligned with available resources. To cover all aspects of a course, some instructors may work around this issue by recommending multiple textbooks. However, instructors find that it can overwhelm students with many resources and financially burden them with the high costs of traditional textbooks.

To overcome challenges posed by traditional textbooks and offer the flexibility instructors seek, we introduce TextCraft – a web application designed to automate custom textbook creation. Drawing from a database of 3 million book chapters sourced from public libraries via a specialized web crawler, our application crafts textbooks tailored to specific course needs. Instructors input their course outline, and the application recommends the ten most relevant textbooks from the database. Through a user-friendly interface, the instructor can preview, select, edit, and merge content from multiple textbooks to produce a customized textbook that aligns seamlessly with their course objectives. This tool transforms the traditionally tedious textbook creation process into an efficient and highly personalized experience.

We evaluate our application based on feedback from instructors teaching a second-year programming course, and we assess how well the output aligns with the input of course outlines. Based on interviews and surveys filled by instructors asking about the usability and relevance of the textbooks created, we got feedback with a high satisfaction rate of over 90%. By incorporating this feedback, we are consistently improving the system. To gauge the effectiveness of our application, we conducted a study assessing the alignment between customized textbooks and the specific requirements of each course. By inputting several course outlines, we assessed whether the matching textbooks adequately covered all the topics. The high matching rate in our initial evaluations highlights the effectiveness of our application in aligning educational content with the specific needs of a course.

Citation
Format

Fan, X., & Timorabadi, H. S., & Emara, S. (2024, June), Board 97: Work-in-Progress: TextCraft: Automated Resource Recommendation for Custom Textbook Creation Paper presented at 2024 ASEE Annual Conference & Exposition, Portland, Oregon. https://peer.asee.org/48398

TY - CPAPER
AB - In the educational landscape, textbooks have long been fundamental pillars for transmitting knowledge. Traditional textbooks often fall short in addressing the unique needs of various courses and the ever-evolving nature of academic disciplines. They lack the flexibility instructors need to tailor materials for specific course requirements or a particular semester. This lack of flexibility makes instructors face the dual challenge of ensuring their course content remains both up-to-date and aligned with available resources. To cover all aspects of a course, some instructors may work around this issue by recommending multiple textbooks. However, instructors find that it can overwhelm students with many resources and financially burden them with the high costs of traditional textbooks.

To overcome challenges posed by traditional textbooks and offer the flexibility instructors seek, we introduce TextCraft – a web application designed to automate custom textbook creation. Drawing from a database of 3 million book chapters sourced from public libraries via a specialized web crawler, our application crafts textbooks tailored to specific course needs. Instructors input their course outline, and the application recommends the ten most relevant textbooks from the database. Through a user-friendly interface, the instructor can preview, select, edit, and merge content from multiple textbooks to produce a customized textbook that aligns seamlessly with their course objectives. This tool transforms the traditionally tedious textbook creation process into an efficient and highly personalized experience.

We evaluate our application based on feedback from instructors teaching a second-year programming course, and we assess how well the output aligns with the input of course outlines. Based on interviews and surveys filled by instructors asking about the usability and relevance of the textbooks created, we got feedback with a high satisfaction rate of over 90%. By incorporating this feedback, we are consistently improving the system. To gauge the effectiveness of our application, we conducted a study assessing the alignment between customized textbooks and the specific requirements of each course. By inputting several course outlines, we assessed whether the matching textbooks adequately covered all the topics. The high matching rate in our initial evaluations highlights the effectiveness of our application in aligning educational content with the specific needs of a course.

AU - Xinyuan Fan
AU - Hamid S Timorabadi P.Eng.
AU - Salma Emara
CY - Portland, Oregon
DA - 2024/06/23
PB - ASEE Conferences
TI - Board 97: Work-in-Progress: TextCraft: Automated Resource Recommendation for Custom Textbook Creation
UR - https://peer.asee.org/48398
ER -