Loading...

Constructing Brand Perceptual Maps from Consumer Reviews of Online Shops Using Machine Learning

Ghadamyari, Mostafa | 2020

369 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 53357 (44)
  4. University: Sharif University of Technology
  5. Department: Management and Economics
  6. Advisor(s): Najmi, Manoochehr
  7. Abstract:
  8. Brand perceptual map is a practical tool for visualizing the position of a brand and its competitors in the mind of customers. In traditional ways of building a brand perceptual map, the researcher identifies important aspects of the product and designs a questionnaire to measure the scores of different brands to gather the required information from users. With the increasing development of online shopping, users have been voluntarily registering their reviews in online shops, in a free and unstructured manner, and have created valuable data sources in these online stores. Due to the large size of registered reviews, processing them requires the use of automated methods in computers. In recent years, machine learning tools in the field of information processing have developed significantly, enabling the processing of large amounts of data with better accuracy than conventional methods. In this study, a method is proposed to create brand perceptual maps from consumers reviews in online stores by using machine learning tools. This method has been applied to tablets in Amazon.com store. It uses natural language processing to identify important aspects of tablets that consumers have discussed the most. Then, the sentiments of the sentences related to these aspects was analyzed using both a machine learning sentiment analysis tool and a dictionary-based tool, and the accuracy of the two methods was compared. The brand perceptual map has been drawn based on the result of sentiment analyses. Among the two methods used to analyze sentiment, it was found that the machine learning-based method performed better than the conventional dictionary-based method
  9. Keywords:
  10. Text Mining ; Sentiment Analysis ; Deep Learning ; Brand Perceptual Map ; Big Data Proccessing ; Natural Language Processing

 Digital Object List

 Bookmark

No TOC