This study leverages the advanced capabilities of the T5 model to delve into Aspect-Based Sentiment Analysis (ABSA) using a unique dataset that elucidates sentiment shifts. Moving beyond traditional sentiment analysis, this research explores the profound changes in audience sentiments triggered by the global pandemic, offering a nuanced perspective on ABSA. By examining the interplay between societal changes and cinematic narratives, this work provides a novel angle on sentiment analysis during transformative events. We compare our proposed model with BERT, LSTM, and RNN, demonstrating that our approach significantly outperforms these models. The T5 model achieves superior performance, with an F1 score of 0.9398, recall of 0.9344, and precision of 0.936, underscoring its efficiency. This study highlights the significant impact of real-world events on the subjective perception of films and showcases T5, along with other advanced models, as comprehensive solutions for analyzing sentiment shifts across different eras.