Analyzing Public Sentiments from Bangladeshi Tour and Travel Vlog Video Comments Using Machine Learning

Public sentiment analysis of tour and travel vlog reviews is crucial in all languages due to its vast usage in the tourism and heritage industry. Unfortunately, to our knowledge, no work has yet been done in this domain for the Bangla language. This study proposes a new dataset of 49,125 reviews collected from 24 different YouTube Bangla tour and travel vlogs which is publicly accessible in the repository https://github.com/cseku170202/Bangla-Tour-and-Travel. The prime goal of this investigation is to employ various machine learning algorithms to precisely recognize the sentiments expressed in Bangla tour and travel vlog reviews and apply explainable AI to provide explanations. Our empirical study considered several algorithms including the ID3, ANN, ResNet, and a Stacking ensemble learner. Among the algorithms implemented, the Stacking model outperformed others and achieved the highest 84.64% accuracy. Local Interpretable Modelagnostic Explanation (LIME) and SHapley Additive exPlanation (SHAP) from explainable AI are also considered in this work to observe the reason behind the model’s performance.