Sentiment Analysis Using Deep Neural Network 1D Convolutional with Long Short Term Memory

Abstract

Both speech recognition and picture recognition have tremendously benefited by deep learning's (DL). Without a doubt, deep learning neural nets have had a significant impact on machine learning (NLP). Due to the enormous amount of information and ideas that are currently produced, shared, and moved via the internet as well as other media, sentiment analysis (SA) is among the most active areas of NLP. This essay's main topic is SA's prevailing NLP problem. Text mining techniques like sentiment analysis are used to find and extract subjective information from text. Sentiment analysis allows businesses to understand how people feel about their brand, product, & service while monitoring online interactions. For sentiment categorization, this research introduces a Hybrid model with Deep Neural Network 1D Convolutional with Long Short Term Memory (1DLSTM). When applied to sentiment analysis, the movie review and amazon review datasets results suggest that the network model can obtain a good classification impact. The preprocessing is applied for text mining, filtering punctuation, and for creating vocabulary. The proposed results were compared with other base models like SVM, KNN, MNB, which shows that the hybrid model outperforms other models.