Abstract:
Recently, the campus of Daffodil International University moved from Dhanmondi to Ashulia.
Consequently, a variety of responses were observed among the students. Based on students' various
reviews to campus changes, a sentiment dataset was created. Sentiment of the students are categorized
into three types e.g., positive, neutral and negative. Different sentiment preprocessing techniques have
been used to preprocess those collected dataset. By the help of this dataset, a machine learning based
Random Forest model is generated to analyze the sentiment of the students. In consequence, this model
can predict any sentiment of a particular student whether it is positive, negative or neutral. Eventually
the accuracy of this sentiment analysis of the students using Random Forest algorithm for shifting the
campus to Ashulia from Dhanmondi is 83.36%. Thus, this model can predict a particular student’s
sentiment with very good accuracy by which authority can take necessary steps to improve the sector
where student’s review was negative.
