IDEAL: an inventive optimized deep ensemble augmented learning framework for opinion mining and sentiment analysis

Sentiment analysis is a method used in machine learning to identify and examine the sentiments that are concealed in text. Annotated data is a requirement for sentiment analysis. This data is frequently manually annotated, which is a laborious, costly, and time-consuming procedure. In this work, a fully automated sentiment analysis annotation method has been devised to overcome these resource constraints. This work develops the clever and novel Inventive Optimized Deep Ensemble Augmented Learning (IDEAL) sentiment analysis system. Cleaning up the social data input is the first step in this data pretreatment process. This includes validation of missing numbers, spelling correction, noise reduction, and standardization. By implementing the Multi-Model Feature Extraction technique, the attributes Word to Vector, Glove, and Bag of Words are recovered from the social data. The ideal subset of features is then chosen using a novel, state-of-the-art technique called the Intelligent Mother Optimization technique (IMOA), which expedites the classifier's training and testing. Furthermore, the classification of attitudes into three categories—positive, negative, and neutral—is accomplished by a classifier model known as Hybrid Convoluted Bi-directional—Long Short Term Memory. The efficacy of the proposed IDEAL framework is evaluated by comparing it to the conventional sentiment prediction techniques and validating a variety of assessment metrics. The overall findings show that, with a 99% efficiency rate and high sentiment prediction accuracy of up to 99.2%, the suggested IDEAL framework performs better than the competition. This is primarily due to the inclusion of novel mining methodologies.