Text mining for the analysis of digital consumer behavior on Twitter

Authors

Plaza, Angel

Marcillo, Julio

Hidalgo, José

Anchundia, Oscar

Pilacuan, Luis

Parrales, Aldo

Navas, William

Abstract

The present work aims to characterize the behavior of the Ecuadorian digital consumer in social networks, for which it is proposed to analyze the comments of the tweets published by Ecuadorian companies that offer telecommunications services. The intense use of social networks by the Ecuadorian consumer, and the change in the business model in Ecuadorian businesses due to the health crisis, requires companies to develop digital marketing strategies, advertising or improve products and services they offer to satisfy the needs of the digital consumer. Within the activities, it is proposed to extract the data through the access tokens of the Twitter API, of the nine processes determined for the treatment of large volumes of data, only six processes that adapt to the requirement of data analysis will be used. In the Jupyter Notebook with the use of Python 3, a word frequency analysis is developed using automatic algorithms. The analyzed results will allow to show positive and negative characteristics of the consumer's interaction in social networks related to the quality of the service.