Enhanced Online product similarity classification using description and Images

Durairaj, Sureshkumar (2023) Enhanced Online product similarity classification using description and Images. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (5MB) | Preview
	PDF (Configuration manual) Download (10MB)

Abstract

Online consumers and retailers share a multitude of data with e-commerce platforms, which is part of the world’s expanding data fission. Magnanimous research has been on the rise especially concerning the detection of product similarities, this invites a clamour towards a possible expanse for research breakthroughs in product similarity identification. Similar products may have modest variances in their textual descriptions on e-commerce sites, but such descriptions may not be as noticeable as the accompanying images . The standard practice among e-commerce platforms in the past was to rely either on text-based comparisons or image-based techniques. The research of fusing title descriptions and image descriptions, however, is expanding as a result of technical developments. In order to identify a list of identical e-commerce products, this research effort uses deep learning algorithms in conjunction with product titles and photos . The method uses a concatenation of ecomBERT+TF-IDFvectorizer for text modelling and ResNet50 v2 with deep layers and ResNet50 for modelling images. Accurately identifying identical products is the aim of experiments involving picture augmentation, embedding, and language processing. The model used in the study, which serves as the basis for evaluation, is based on a mix of eComBERT and ResNet50v2. Cross-validation scores are used to gauge the correctness of the model, while computing time is used to gauge efficiency . The results of this study will analyze the implementation results and include a comparison study utilizing actual data from a well-known e-commerce business such as shopee.com.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Ul Ain, Qurrat UNSPECIFIED
Uncontrolled Keywords:	product similarity match; e-Commerce; TF-IDF; e-ComBERT; ResNet-50v2
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Algebra > Algorithms > Computer algorithms H Social Sciences > HF Commerce > Electronic Commerce Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	22 Nov 2024 10:47
Last Modified:	22 Nov 2024 10:47
URI:	https://norma.ncirl.ie/id/eprint/7183

Actions (login required)

View Item