Twitter(X) Sentiment Analysis Using Neural Ne...

Twitter(X) Sentiment Analysis Using Neural Networks

Author Name : Shivaprakash D M, Asst Prof. Uma Mageswari

ABSTRACT Sentiment analysis has been a outstanding cognizance in the discipline of language knowledge, even though there remains room for development inside the software of neural networks. Currently, most studies emphasizes sentiment detection by using reading syntax and vocabulary. Moreover, obligations associated with herbal language processing (NLP) and attaining top notch results have employed Recurrent Neural Networks (RNNs) and Convolution Neural Networks (CNNs). CNNs, however, require the usage of more than one layers to effectively seize lengthy-term dependencies. In this paper, we suggest a hybrid structure that first utilizes RNNs to seize long term dependencies, observed by means of CNNs with a global average pooling layer. Additionally, we comprise a phrase embedding approach the use of GloVe(Global Vectors for Word Representation),trained through unsupervised gaining knowledge of on a big Twitter corpus to cope with this difficulty.