| dc.description.abstract | Customer satisfaction is one of the most important scales of how good a hotel is.
Sentiment analysis on short sentences on the Internet has gotten a lot of interest
because of the expanding volume of text online and the value it gives to society in
general and hotel service in particular. In this thesis, the purpose of analyzing textual
data is for reviewing the comments of the customers when they have just used the
service. A BERT model acknowledged Transformer is brought up in order to indicate
the sentiment element presented in customer expression. A hotel in Mussoorie, which
is a beautiful place, is chosen to perform in this thesis. The owner of this hotel is an old
guy who wants to predict the thinking from the customers’ perspective. And results
from the calculation are very significant and promising for use in other data. The
prediction algorithm gives out a highly efficient accuracy. This approach's
performance is quite effective, and it processes textual datasets relatively well. Despite
the high quality of the data analysis, this model is not the best option for every
sentiment analytics situation. Because several languages have distinct writing styles,
extra strategies for data evaluation will be required. | en_US |
