A Smart Framework to Analyze Hotel Services after COVID-19

ABSTRACT

In the last decade, many websites have allowed customers to provide reviews about their experiences in dealing with different hotels like describing their opinions about different sides like cleanliness, food, employees, and other services. However, after COVID-19, new concerns have emerged linked to the health precautionary and preventive processes. At the same time, the amount of generated data (Textual Data) has become very huge (Big Data), and it needs to auto classifier for processing it where it is impossible to review manually. So, this research proposed a smart method to detect useful information from text by extracting the interesting services from comments automatically. The method depended on machine learning (ML) and compared between five different classification models (Spacy, Naïve Bayes (NB), Stochastic Gradient Descent (SGD), Passive Aggressive, and AdaBoost) to determine which one provides the best results and accuracy. According to the experiment on a real dataset containing more than 1000 records, the NB was the best accurate with 98%, while the Spacy was less and that related to the small size of training data. © 2022 Bharati Vidyapeeth, New Delhi.