This study aims to examine students’ opinions, sentiments, assessments, attitudes, and emotions in written language. Sentiment analysis is one of the most active study fields in natural language processing, as well as a popular topic in data and text mining. In reality, we now have digital access to an enormous amount of philosophical data that can be evaluated. It is known as sentiment analysis and opinion mining and is faced by students taking the flexible learning modality during a baccalaureate degree program administered by Eastern Samar State University with the use of Sentiment Reasoning through Lexicon-Based approaches. This approach can be further classified into dictionary-based approaches and corpus-based approaches. The researcher analyzed and visualized a corpus of texts, including student opinions and attitudes, using machine learning technology. To examine the students’ feelings, Lexicon analysis was applied, and Latent semantic indexing topic modeling was used to build topics from the corpus of papers. . The data was gathered from the College of Computer Studies Students (n=403) of Eastern Samar State University during the school year 2020-2021. The collected data is then translated to a digital format and compiled into a single data set. It was observed that the majority of students had a negative or undesirable attitude about the use of flexible learning as a mode of instruction, according to the findings. Themes that were revealed as a result of the compilation were put into a meaningful structure to be presented to the reader. The results showed that the following problems occurred in the process of the students have when faced with flexible learning, according to the topic modeling. It was also discovered that some students were employing survival techniques and experimenting with novel learning practices in order to complete the tasks that had been assigned to them. The main outcome of this study was that the need for an orientation program which will assist students in taking flexible learning modalities.