Identifying fake users using an integrated system using NLP

Ramachandran, Karthik (2023) Identifying fake users using an integrated system using NLP. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (3MB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

Fake reviews on online platforms have emerged as a alarming issue. Consumer perceptions on any product is influenced by deceptive or biased feedback posted in online platforms. The integrity of product or service ratings can be distorted by these fake reviews, which can also undermine user confidence in buying the product. This study proposed a novel approach on detecting users who post fake reviews with an integrated system using embedding vectors and clustering similar reviews using cosine distance. A Long Short-Term Memory (LSTM) model was built to predict the user ratings and this model is used to transform each review to its own embedding vectors. Grouping of similar embedding vectors is achieved by finding cosine angle between the vectors. Google’s ScaNN was implemented to calculate the cosine distance between the vectors and for faster search and retrieval. The model was built with an accuracy of 78% for test data and the grouping of similar reviews were achieved.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Hafeez, Taimur UNSPECIFIED
Subjects:	H Social Sciences > HF Commerce Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing H Social Sciences > HF Commerce > Marketing > Consumer Behaviour
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	28 Dec 2024 15:20
Last Modified:	28 Dec 2024 15:20
URI:	https://norma.ncirl.ie/id/eprint/7255

Actions (login required)

View Item