Bridging Qualitative and Quantitative Methods: LDA and Sentiment Analysis in Computational Communication

Technological advancements have transformed recent communication research, introducing computational methods like Latent Dirichlet Allocation (LDA) for topic modeling and sentiment analysis for affective evaluation. However, these tools face limitations, including contextual sensitivity, evolving language use, and restricted interpretability. This research mainly used case studies to present the tools and results of previous cases, comparing their advantages and disadvantages. As a result, current studies often overlook multi-agent interactions and narrow analytical scopes. Thus, future research should integrate qualitative insights with quantitative data, adopt diverse theoretical frameworks, and improve algorithmic transparency. Emphasizing multi-agent dynamics will deepen analytical rigor and extend the applicability of computational approaches in complex digital communication environments.