摘要:Text content on social platform is a good source to understand the human opinion for a brand, organization, product, service, etc. Opinion mining from digital micro blogs help analyst to understand, resolve different issues. This paper contributes to improve the opinion class detection accuracy of the work by involving a hybrid combination of genetic algorithm with neural network. Genetic algorithm Butterfly Particle Swarm Optimization was used to cluster patterns from the input training micro blogs. Cluster patterns act as input training feature vector while opinion class act as desired output for training of neural network. This trained neural network predicts the opinion class of testing micro blogs dataset. Experiment was done on twitter microblog dataset. Results show that proposed GANNOM (Genetic Algorithm and Neural Network Based Opinion Mining) has increased various evaluation parameters of the work.