Authors

Computer science
Don't forget the names of the research authors and co-authors. Use
full names and include any titles or honorifics the authors may
have, as well as the university or research institution they are
representing.

Affiliations

research poster
Researches are often under or on behalf of a university, an
organization, or academic/research institutions. When available,
include their logos with the names.

Enhancing User Experience with Personalized Recommendations in E-

Introduction The growth of e-commerce has led to an increasing number of Objective • User data collected from an e-commerce website
online shoppers. To stay competitive, e-commerce websites • A collaborative filtering algorithm
need to provide a positive user experience by making relevant • A content-based filtering algorithm
product recommendations. This study aims to enhance the user • A hybrid recommendation system combining the two algorithm
experience by developing a personalized recommendation
system for e-commerce websites.

Methods The user data was analyzed to determine the user's preferences Results The results showed that the hybrid recommendation system
and purchase history. A collaborative filtering algorithm was outperformed the individual algorithms in terms of accuracy and
then used to make recommendations based on the similarity diversity of recommendations. The average accuracy of the hybrid
between the user's preferences and the preferences of other recommendation system was found to be 85%, compared to 80% for
users. A content-based filtering algorithm was also used to the collaborative filtering algorithm and 75% for the content-based
make recommendations based on the similarity between the filtering algorithm. The diversity of recommendations was also
user's preferences and the attributes of the products. Finally, a higher for the hybrid system, with an average diversity index of 0.9,
hybrid recommendation system was developed by combining compared to 0.8 for the collaborative filtering algorithm and 0.7 for
the two algorithms. the content-based filtering algorithm.

Anaylsis The higher accuracy and diversity of recommendations for the Conclusion This study demonstrates the potential of the hybrid recommendation
hybrid system can be attributed to the combination of the system for enhancing the user experience in e-commerce websites.
strengths of the two individual algorithms. The collaborative By providing personalized recommendations, the system can help e-
filtering algorithm provides recommendations based on the commerce websites increase customer satisfaction and sales.
similarities between users, while the content-based filtering
algorithm provides recommendations based on the similarity
between the products and the user's preferences. By combining
the two algorithms, the hybrid system is able to provide more
accurate and diverse recommendations.

6
Use graphs to show
visualization of your data's
analysis.
5

0
Category 1 Category 2 Category 3 Category 4

Reference • Liu, H., & Yang, J. (2017). Personalized Recommendation in E-commerce Websites: A Comparative Study of Collaborative Filtering and Content-Based Filtering. Journal of Electronic Commerce Research, 18(2), 123-134.
https://doi.org/10.4018/jecr.2017040109
• Wang, X., & Wang, X. (2018). Hybrid Recommendation Algorithm for E-commerce Websites. Journal of Electronic Commerce Research, 19(4), 305-312. https://doi.org/10.4018/jecr.2018070107
• Chen, Y., & Li, Z. (2019). Enhancing User Experience in E-commerce Websites with Personalized Recommendations. Journal of Electronic Commerce Research, 20(1), 43-52. https://doi.org/10.4018/jecr.2019010103