Enhanced Convolutional Neural Network based Sentimental Analysis Framework for Health Care Domain

  • D. Sasikala, Dr. S. Sukumaran

Abstract

Sentimental Analysis (SA) is growing rapidly across various domains as a direct outcome of natural language processing and machine learning approaches for assessing and classifying emotions from subjective data.  Analysing sentiment over healthcare domain is the cumbersome task, which provides a variety of information regarding patient lifestyle. Hence high level of impedance involved while classifying the sentiment from the clinical documents. Existing researches have been focused on summarization of  text, reduction of features, and prediction of sentiment separately.  In this research work, all the approaches are integrated to provide a novel sentimental analysis framework for classifying emotions of a patient.  The proposed work is consisting of three folds.  Initially, pre-processing is done which includes tokenization, stemming, lemmatization, stop words, lower case conversion.  First, text summarization is performed, which utilises word2vec with condition random field, Second, Selection of features is done with help of Improved Genetic Algorithm (IMGA), hence as a result standard features only exist which determines fittest individuals.  Finally, the accurate sentiment prediction is done with help of Enhanced Convolution Neural Network (ECNN).  The proposed ECNN incorporate 1D layer, LSTM layer and regression layer. The proposed work is experimented by utilising the two real-time datasets, which is widely consisting of medical reviews from social media platforms. To analyse the proposed patient sentimental analysis model, the evaluation metrics like accuracy, precision, recall and f-score are used and, also comparison has been made with existing state-of-art model. our proposed work outperforms well than other methods in terms of all performance metrics.

Published
2021-11-22
How to Cite
D. Sasikala, Dr. S. Sukumaran. (2021). Enhanced Convolutional Neural Network based Sentimental Analysis Framework for Health Care Domain. Design Engineering, 14897-14908. Retrieved from http://thedesignengineering.com/index.php/DE/article/view/6614
Section
Articles