Comparative Analysis of Multi-Model and Uni-Model Approaches using Time Distributed Bidirectional LSTM for Multidata Sentiment Analysis
Krati Gupta, Mahesh Parmar · 2023
The sentiment of the data contained in a text can be determined by analysis. This sentiment can be good, negative, or neutral. When carrying out this research, one may choose to employ either a uni-model or a multi-model technique. In the uni-model method, both the training and the analysis are performed using just a single model. The application of this strategy is uncomplicated and uncomplicated, but it might not be able to make sense of complicated or different facts, which would lead to wrong findings. Instead, the multi-model technique utilizes a number of different models, each of which is designed to examine a particular aspect of the data or a particular kind of attitude. This allows the multi-model strategy to produce more accurate results. The results of several different models might be combined in order to obtain a more in-depth understanding of feelings. This method is more accurate overall and captures more complex expressions of emotion, but it requires more time and money to train personnel and put into practice. The evaluation of metrics was carried out using Multimodal and mono Model with Time Distributed Bidirectional LSTM in this research. Both the loss and the accuracy. The Multi-Model achieves the highest accuracy of 99.71 percent for the train, while the validation accuracy achieves 99.32 percent, and the Multi-Model achieves the least loss of 0.010 percent.