Key points are not available for this paper at this time.
In the context of existing sentiment analysis, comments are classified as positive, negative, or neutral. In some cases, sentimental comments contain a dual meaning within a single sentence. However, the sentence also includes both positive and negative vibes. In such cases, negative comments are classified as neutral or ignored (for example, the dress appears to be really nice to wear, but it's worth the price because it's simple and there's no heavy work). Based on TF-IDF technique, this research study proposes a fuzzy based weighted sentiment classification by using transformer neural networks. Based on the data set, the system will analyze both the positivity and the negativity of the sentences by calculating their weights. In the proposed system, the weighted value is used to classify positives and negatives with a minimum threshold level. If the calculated weights are the same mean, then the system is considered neutral. This system enables to find out the average weighted factor of positive comments for product promotion. As an alternative, the seller can indirectly improve quality by specifying negative average weighted factors. Overall, the system was evaluated in terms of Precision & Accuracy and produced 95.6% & 96% respectively, and its results were better than those of other existing systems.
Shrivastava et al. (Thu,) studied this question.