Gaining Insights from EV Charging Reviews Using Natural Language Processing

Quaicoe, Hector (2022) Gaining Insights from EV Charging Reviews Using Natural Language Processing. Bachelor's Thesis, Computing Science.

Preview	Text Thesis.pdf Download (669kB) | Preview
	Text toestemming.pdf Restricted to Registered users only Download (126kB)

Abstract

Customer satisfaction plays an imperative role in the business' success. One way to measure customer satisfaction level is by utilizing customer reviews. This thesis analyzes customer reviews of EV charging stations owned by Shell using a text mining approach, including sentiment analysis and topic modeling. Vader Lexicon is the classification method utilized to aggregate positive or negative sentiments in each review. Moreover, Latent Dirichlet Allocation is used to cluster reviews into various topics. In addition, a Support vector machine classifier is used to identify positive or negative sentiments in the review sentence to compare the unsupervised and supervised approaches taken. The classification results show that the supervised technique performs better at classifying sentiments. Ultimately, this thesis aims to help Shell use their data efficiently by improving the quality of its EV charging solutions and staying ahead of its competitors.

Item Type:	Thesis (Bachelor's Thesis)
Supervisor name:	Degeler, V. and Tello Guerrero, M.A.
Degree programme:	Computing Science
Thesis type:	Bachelor's Thesis
Language:	English
Date Deposited:	08 Jul 2022 13:32
Last Modified:	08 Jul 2022 13:32
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/27678

Actions (login required)

View Item