Proposing a Sustainable AI Assistant to Reduce Environmental and Economic Cost of Fast Fashion

Wonjin Eum
Stevenson School

Abstract

Fast fashion has shifted modern consumer culture by promoting overconsumption and inconsiderate purchases, leading our team to challenge ourselves to fight against this issue. The overbearing environmental, economic, and ethical problems that fast fashion has brought to society are becoming detrimental, both individually and globally.
This AI assistant fosters sustainable consumer culture by utilizing machine learning and patterns.
Incorporating the neural network system, creating a unique score function, and adopting recommendation systems, the assistant ultimately provides users with a “purchase score” that helps them minimize careless purchases.
This filtering model is generated using a deep learning code with a refined dataset. Then, the user input will be passed onto the model to get the prediction. Based on this information, the purchase score will come up via the Score function I defined. With the Score function, the Recommendation system would provide recommendation indicators by user-based and item-based systems. Finally, the optimization and personalization will present a sophisticated analysis of the consumer’s purchase patterns.
The AI-assistant implements Python to build the machine learning model, and the pseudo-code and different parameters are included in the main text. The research puts emphasis on finding ways to increase the efficiency of the process and concludes with a suggestion on pattern analysis using deep learning, and vision training for the user to upload their history easily.

Please click to read the paper.