Enhancing Tailored Discovery on TikTok Shop

Inspiration:

The biggest inspiration for this project is derived from our real experience with TikTok Shop. As new users to TikTok Shop, we love how it continuously recommends products and pushes the recommendation on ads whenever we scroll TikTok videos. However, we found some drawbacks whenever we try to find new products on TikTok Shop as it does not recommend the products the we desire. We would love to address a solution to elevate the recommendation system on TikTok Shop to bring more personalized shopping experience.

What it does:

The recommendation system uses Hybrid recommendation system to pushes the top rated products and recommend them to users based on the user's profiles, past activities, purchase history, etc. Even for new users, we have some categories for them to choose and recommend top 10 popular products for them.

How we built it:

We used Google Colab Notebook to represent the algorithm and Figma to illustrate the experience.

Challenges we ran into:

In undertaking this project, our team encountered several significant challenges. The first major issue we faced was determining how to effectively collect data for our recommendation system algorithm. Gathering accurate and comprehensive data is crucial for creating reliable recommendations, and finding the right approach was a complex task. Additionally, we struggled with illustrating our backend and frontend processes clearly. Ensuring that both the technical and user-facing aspects of the project were well-integrated and visually coherent required considerable effort and collaboration. Finally, managing these tasks while adhering to tight deadlines added an extra layer of difficulty, pushing our team to constantly refine our strategies and work efficiently.

Accomplishments that we're proud of:

Even though we did not manage to build an application on time, we are happy with the effort that we put in this project.

What we learned:

In this project, we learned the importance of collecting accurate data and cleaning it throughly to ensure our recommendation algorithm work well. We tried different methods and focuced on using popularity-based suggestions for new users and personalized recommendations for returning users, helping us tailor our approach to different types of users. Working on intergrating the backend and frontend taught us the value of good teawork between technical and design teams. We had to communicate clearly to createa seamless user interface. Managing tight deadlines improved our project management and prioritization skills, showing us how to keep quality high even when time is limited. These experiences in data handling, developing algorithms, collaborating as a team, and managing time will be very useful as we keep improving the TikTok Shop to make it more engaging and user-friendly for everyone.