Product Matching for E-commerce Platform based on Text and Image Similarity using Deep Neural Network Architecture

Sanaullah Shariff, Zahra Fathima (2022) Product Matching for E-commerce Platform based on Text and Image Similarity using Deep Neural Network Architecture. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (3MB) | Preview
Preview	PDF (Configuration manual) Download (705kB) | Preview

Abstract

Many businesses are concentrating on their e-commerce nowadays since it helps businesses sell their goods over a wider range and makes it easier to identify their clients and their needs. We all experience situations like taking time to locate favourites products and waiting in line at the register in a store on a regular basis. Also over the online platforms, many sellers are selling similar product by adding the different description of product, tags and titles. In such situation, it becomes very difficult for the e-commerce players to identify the identical products. Therefore, identification of similar product is a necessary task and still an open area of research also automation is another requirement for the advancement of society and the economy since it saves time while also being more dependable than manual processes. Deep Learning has made enormous strides recently, including successes in object detection and image categorization. Matching similar products based on the input image and text which may be tag or label can be recognized by utilizing the convolutional neural network. In this research work, three pre-trained deep convolutional models (MobileNet-V2, VGG-19, ResNet-50) are executed on image and text data to predict the most similar images with each model. Cosine similarity, Levenshtein distance, and custom metric score are some of the performance measure technique has been utilized to identify the most optimal model for similarity detection based on text and image. According to the study, the Mobilenet model is the finest ideal model for dealing with such time-consuming chores and aiding in growing a digital strategic plan.

Item Type:	Thesis (Masters)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science 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:	10 Mar 2023 16:35
Last Modified:	10 Mar 2023 16:35
URI:	https://norma.ncirl.ie/id/eprint/6292

Actions (login required)

View Item