Multi-Class Sentiment Classification for Healthcare Tweets Using Supervised Learning Techniques

Multi-Class Sentiment Classification for Healthcare Tweets Using Supervised Learning Techniques

Brahami Menaouer, Abdeldjouad Fatma Zahra, Sabri Mohammed
DOI: 10.4018/IJSSMET.298669
Article PDF Download
Open access articles are freely available for download


Social media has revolutionized the way people disclose their personal health concerns and express opinions on public health issues. In this paper a new approach for multi-class sentiment classification using supervised learning techniques. The aim of this multi-class sentiment classification is to assign the healthcare Tweets automatically into predetermined categories on the basis of their linguistic characteristics, their contents, and some of the words that characterize each category from the others. Briefly, relevant health datasets are collected from Twitter using Twitter API; then, use of the methodology is illustrated and evaluated against one with only three different algorithms was used, to improve the accuracy of Decision Trees, SMO, and K-NN classifiers. Many experiments accomplished to prove the validity and efficiency of the approach using datasets tweets and it accomplished the data reduction process to achieve considerable size reduction with the preservation of significant dataset's attributes
Article Preview

1. Introduction

Today, Social Network Analysis (SNA) is commonly applied to investigating trends with studies of Social Media Analysis (SMA) and Data Mining (DM) used for such purposes (Khalifa et al. 2021; Wahi et al. 2014). In the last decade, Twitter has emerged as the most influential micro-blog service with twitter data source gaining considerable attention among researchers. Unlike many other social network services, twitter makes most user data world accessible. Additionally, twitter has numerous amounts of tweets, which mainly express opinions about a diversity of topics. These tweets may express valuable feedbacks and attitudes from patients about a specific disease or medical treatments. Twitter, unlike many other social networking services, makes most user data publicly accessible. These tweets, which express opinions about diverse topics, may also offer valuable feedback and reflect insightful attitudes from patients on specific diseases or medical treatments. O’leary (2015) observes that Twitter has become a critical social media tool with key capabilities, such as communication, building communities and collective action organizations. Yet, the fact that texts in social media are mostly written in colloquial language and both understanding and analyzing these texts is somewhat difficult in medical (epidemiological) context, further research attention in this area is needed, especially in text sentiment analysis. Specifically, sentiment analysis examines how sentiment is expressed in texts; briefly, sentiment systems may be used to identify sentiment categories from texts (Gridach et al. 2018; Tariyal et al. 2018).

In an earlier work, sentiment analysis is an automatic analysis technology of written or spoken speeches, aiming to extract subjective information like judgments, evaluations or emotions to detect the polarity of an opinion. According to (Kaur & Kautish, 2019; Bansal & Kaur, 2018; Rathi et al. 2018), sentiment analysis can be defined as the process of categorizing the opinions expressed through tweet to understand the user views about that topic. For (Yasen & Tedmori, 2019), sentiment analysis has been proposed as a component of other technologies. One idea is to improve information mining in document analysis by excluding the most subjective section of a text or to automatically propose internet ads for products that fit the viewer’s opinion and removing the others. In general, sentiment analysis is performed utilizing one of two main approaches: the lexicon-based or unsupervised learning approach, in which rules extracted from the linguistic study of a language are applied to the sentiment analysis and the machine learning or supervised learning approach which relies on the famous machine learning algorithms to solve sentiment analysis as a classification task. Besides, the techniques for sentiment analysis on Twitter have been categorized into Lexicon Based, Machine Learning Based, Hybrid (Lexicon+Machine Learning) and Concept based (Ontology or Context) across pertinent literature (Kumar & Jaiswal, 2017).

Complete Article List

Search this Journal:
Volume 15: 1 Issue (2024)
Volume 14: 1 Issue (2023)
Volume 13: 6 Issues (2022): 2 Released, 4 Forthcoming
Volume 12: 6 Issues (2021)
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing