![]() If irony detection fails, the performance of sentiment analysis may suffer. ![]() The existence of irony may produce polarity reversal. If the predefined meaning is the inverse of what it transmits, the text word is seen to be ironic. Figurative and creative language, such as sarcasm and irony, can be found in user-created comments on social networks. In the last twenty years, there has been considerable progress in analysing sentiment in English. Sentiment analysis is also used to analyse online user-created content in real-world settings such as business-based activities, product evaluations, and opinion mining analysis. Sentiment analysis is the most important and renowned task for automatically estimating the sentiment of online user-created text. Investigating text-based social network content is more important for gathering data in the field of scientific research and management (Chen et al., 2020). The social network uses sentiments, intentions, emotions, and views to share data. These networks commonly use natural language processing (Li et al., 2020). These communication networks generate a tremendous amount of information every day, which grows into a perceptive data source for many research purposes. Social networks like Facebook and Twitter are mostly utilised by individuals for openly sharing their thoughts because to the recent popularity of a web 2.0 appliance (Ahmad et al., 2019). The simulation results obtained by comparing the proposed model with the various existing techniques display the effectiveness in terms of F1 score, precision, classification accuracy, error rate prediction, and kappa statistics. The ARO-optimized RRVFLN model classifies the outcomes obtained in the three datasets namely the News Headline Sentiment dataset, Sentiment140 dataset, and 4000 Short Story dataset into positive and negative. Artificial Rabbits Optimization (ARO) is utilized to RRVFLN model’s input weights and to optimize the hidden layer biases. Initially, the data is preprocessed and the word vectors are efficiently analyzed by the Fast text method by considering each token as an n-gram character. Motivated by this problem, this paper presents an Artificial Rabbits optimized Robust Random Vector Functional Link Network (RRVFLN) for improving sentiment analysis accuracy. However, very few researchers have used the randomization-based neural network for sentimental analysis which offers different benefits such as increased generalization capacity, improved accuracy, low training time, and universal approximator for non-linear functions. However, the different complexities associated with deep learning techniques such as high computational and space complexity, sensitivity to learning rate, high training time, etc make them hard to implement in real-time applications. ![]() The deep learning techniques combined with Natural Language Processing (NLP) has said to offer higher performance in analyzing the sentiments in implicit aspects present in high-quality training datasets. Sentiment analysis using artificial intelligence is a great tool for analyzing human linguistic capabilities.
0 Comments
Leave a Reply. |