An Attribute Selection Using Propagation-Based Neural Networks with an Improved Cuckoo-Search Algorithm

Sentiment Analysis is getting an area of implication of the researchers in the business as well as in the research. To know about the opinions of human beings through artificial machines have always been interesting thing to note and that has been getting updated from all over the world as the time has passed by. The research article presents an attribute selection mechanism by using enhanced Cuckoo Search algorithm which is known as meta-heuristic categorized algorithm. A novel fitness function which has been designed and Neural Networks have been used for the training and validation to get the proposed solution. The proposed algorithm has also been compared to state art of the art techniques based on quantitative parameters. Accuracy has been considered as a main objective. The detailed result and analysis have shown that integration of neural network have demonstrated distinguishing results in terms of performance parameters, namely, precision, recall, f-measure, and accuracy of polarity classification.