Thematic Analysis of Students’ Perception and Attitudes towards Online Class using Latent Dirichlet Allocation

The Latent Dirichlet Allocation (LDA) topic model is employed for the purpose of categorizing textual content within a document into specific topics. The objective of this study is to construct a model that identifies themes related to students’ perceptions and attitudes towards the deployment of online classes. This was achieved by analyzing data collected from students at CSPC who are now experiencing the effects of the epidemic. The participants of the study were restricted to those who were currently enrolled as students in Polytechnic Colleges located in the province of Camarines Sur. Data collection was conducted through the utilization of questionnaires, which were distributed to the students via Google Forms. The present study employed a qualitative research design. The efficacy of the LDA algorithm in topic or theme identification was observed by the researchers. The utilization of coherence score and perplexity proved to be a dependable approach for testing the model. The visual depiction of the themes effectively highlights the significant keywords associated with each issue and yields favourable outcomes that accurately portray the extracted topics about the sentiments of the pupils. The selection of the topic name and the subsequent interpretation effectively elucidated the significance of each respective issue.