Table 1.

Summary of existing sentiment analysis studies.

S.No	Reference	Technique	Advantage	Limitation
1	[14]	MaxEnt-JABST	Efficient for opinion extraction	Possess cross-domain sentiment analysis issues
2	[15]	NB and MaxEnt models	Carry out sentiment analysis at abstract levels	Less accuracy
3	[18]	DBNSA	Effectively classification depending on user ratings	Computationally complex
4	[19]	Neural network model	Efficient for binary sentiment classification	Extracted features do not suit nonbinary sentiment classification tasks
5	[21]	AT-MC-BiGRU-capsule	Used a capsule mechanism for text characterization	Stability issues
6	[24]	RNN	Effective even for larger data	Glove feature extraction resulted in lower accuracy
7	[25]	Deep CNN-LSTM	Efficient for review analysis in the e-commerce domain	Requires higher computational power
8	[30]	Optimized dictionary-based multilabel classification	Dynamically categorizes nonfunctional needs from app user feedback	Less efficiency
9	[33]	RF, SVM, and NB	Analyzes textual reviews of digital payment apps	Cost-inefficient
10	[29]	CNN	Effective classification of app reviews	Feature extraction is not efficient