Project META Overview

Project Objective

Project META aimed to create a metaverse platform for the client UPPURPLE, a company that provides a solution for problems embedded in online shopping malls through a multi-complex metaverse fashion shopping platform. The main objective of the project was to create a shopping mall application with a metaverse shopping mall platform and avatars customized for users of various types.

Team Objective

The Data Science team played a crucial role in the project by programming a chatbot to help navigate users through the online shopping mall, developing a recommendation system that provides recommendations to customers based on their preferences, and developing various interactive dashboards where customers and shop owners can observe various datasets including user activity and purchase history.

<aside> 💫 Team Goal(s):

To program a chatbot to help customers navigate the online shopping mall better and receive help without the need for human interaction
To develop a recommendation system where customers can get recommendations of clothing based on the image that they upload
To develop interactive dashboards that visualize customer information and behavior based on the information gathered from chatbot responses, shopping mall activity, and product purchase history for managers to devise customer-targeted ads/events </aside>

About Client

UPPURPLE is a multi-complex metaverse fashion shopping platform that aims to demolish the boundary between offline and online shopping for consumers and retail shops. Initially a South Korean startup funded by the government under the U Tech category, UPPURPLE made a strategic decision in November to prioritize the U.S. market due to the higher adaptability of the metaverse concept there. The company was acquired by Divelabs, a tech firm based in New York known for its involvement in new technologies like NFTs and Web3. The project continues under the name UPPURPLE within Divelabs, maintaining its identity despite being absorbed by its parent company.

2023 Fall Capstone

July 24th - Dec 18th

Topic Highlights

Basics of Coding

Data Visualization

Natural Language Processing (NLP)

Web Development

Web-based Dashboard

Recommender Algorithms and Systems

Movielens Dataset Application

Analyzing Purchase History Using Apriori Algorithm

Clothing Recommendation System

Deliverables

Purchase History Dashboard

Clothes Recommendation Dashboard

META Web Dashboard

Next Steps

Chatbot Enhancements: Continue refining the chatbot's capabilities to offer more seamless and comprehensive assistance to users navigating the online shopping mall. Incorporate machine learning and natural language processing advancements to improve the bot's understanding and responsiveness to diverse user queries.
Advanced Recommendation Algorithms: Explore and implement advanced algorithms for the recommendation system. Utilize deep learning models or image recognition techniques to enhance the system's accuracy in suggesting clothing items based on uploaded images, ensuring more precise and personalized recommendations.
Dashboard Evolution: Further develop the interactive dashboards to provide more insightful visualizations and analytics. Expand data integration to include real-time updates on user behaviors, preferences, and trends within the shopping mall. Consider predictive analytics to anticipate customer interests and enhance targeted marketing strategies.
Experimentation and Adaptation: Engage in continuous experimentation with emerging data science technologies. Stay updated on the latest advancements in AI, machine learning, and data visualization to adapt and implement innovative features within the META platform, ensuring it remains at the forefront of user experience and data-driven decision-making.
Collaboration and User Feedback: Collaborate with other teams and departments to leverage diverse insights and data sources. Actively gather user feedback and analyze user interactions with the chatbot, recommendation system, and dashboards to implement improvements and tailor features according to user preferences.

Untitled