User-Driven Product Exploration in E-Commerce Using Generative AI and Explainable NLP
Jayawardena, Thishan (2024)
2024
Ohjelmistokehityksen maisteriohjelma - Master’s Programme in Software Development
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2024-11-20
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-2024111110090
https://urn.fi/URN:NBN:fi:tuni-2024111110090
Tiivistelmä
The characteristics of e-commerce are at the stage of constant development and improvement, so it is important to research and find ways to improve the current model for customers and their needs. The current methods used for search and recommendation such as keyword-based search and recommendation do not fully meet the needs of users since they fail to capture the complexity of users’ preferences, and there is no accountability or explainability of the results. This work seeks to investigate the possibility of using Generative AI coupled with Explainable NLP to enable explorative product search by the user on an e-commerce site. Explainable NLP interfaces of the proposed system enable users to type in a description of the products they require in normal language. From this understanding, the Generative AI offers the related products or generates the new copies based on the user’s description.
It is an undeniable fact that technologies that make up the NLP tool should have the capacity to offer transparency so the users can discover how the inputs provided are analysed and how the recommendations are arrived at are aspects that explainable NLP techniques cover. It forms confidence in the system and the decision made and in turn, the satisfaction of the users is boosted. Computer generated product variations with the help of generative AI resources like GANs and VAEs are more suited to help in the discovering of products and making the products more personalized.
It also focuses on the mechanism-based issues as well as the ethical concerns that arise in the form of recommendation systems based on big data and AI technology, how to reduce the prejudice and recognition of the user’s information. Therefore, through the combination of Generative AI and Explainable NLP methods in this study, the vision is to enhance the customer’s experience and achieve a more enjoyable, hence, a user-oriented e-shopping experience for the targeted online store. The provided comprehensive approach creates the background for brand-new, personally oriented, and trustworthy e-commerce interaction with the client, increasing their loyalty.
It is an undeniable fact that technologies that make up the NLP tool should have the capacity to offer transparency so the users can discover how the inputs provided are analysed and how the recommendations are arrived at are aspects that explainable NLP techniques cover. It forms confidence in the system and the decision made and in turn, the satisfaction of the users is boosted. Computer generated product variations with the help of generative AI resources like GANs and VAEs are more suited to help in the discovering of products and making the products more personalized.
It also focuses on the mechanism-based issues as well as the ethical concerns that arise in the form of recommendation systems based on big data and AI technology, how to reduce the prejudice and recognition of the user’s information. Therefore, through the combination of Generative AI and Explainable NLP methods in this study, the vision is to enhance the customer’s experience and achieve a more enjoyable, hence, a user-oriented e-shopping experience for the targeted online store. The provided comprehensive approach creates the background for brand-new, personally oriented, and trustworthy e-commerce interaction with the client, increasing their loyalty.