Sentimental analysis of social networks: A comprehensive review (2018-2023)

Abstract

The field of sentiment analysis, a burgeoning field dedicated to deciphering emotions, opinions, and attitudes within text data, has experienced remarkable growth in research interest. This review paper embarks on a comprehensive journey through the dynamic landscape of sentiment analysis, encapsulating its multifaceted dimensions across three prominent academic databases—Scopus, the IEEE Xplore Digital Library, and the Wiley Online Library—spanning from 2018 to 2023. This paper establish the significance of sentiment analysis, charting its evolution from 2008 and elucidating its multifarious applications across domains. We underscore the necessity of drawing insights from diverse databases, transcending the confines of a single source. "On the causality of temporal sentiment analysis," delves into the intricate interplay between sentiment signals and other variables, echoing Granger causality theory from econometrics. We unveil the burgeoning adoption of Granger causality in examining stock market dynamics and sentiment analysis, shedding light on the predictive causality landscape. The temporal dimension takes centre stage "On the temporal dynamics of sentiment analysis," where we explore how sentiment fluctuates over time, particularly during transformative events such as the COVID-19 pandemic. We scrutinize the evolution of emotional appeals in political campaigns and the pivotal role of sentiment analysis in detecting online bullying and suicidal ideation. "On the reproducibility and practical application," we deliberate on the challenges and prospects of implementing sentiment analysis techniques in real-world scenarios. We navigate the intricate landscape of reproducibility and translate research insights into actionable strategies."Discussion and conclusion," encapsulates the essence of our journey. We highlight content variations across databases and unravel the evolution of techniques. We contemplate the bridge between academia and industry by presenting an integrated view of sentiment analysis. As the sentiment analysis field evolves, this review paper offers a panoramic view of its past, present, and future. It is a valuable resource for researchers and practitioners seeking to navigate the ever-expanding horizons of sentiment analysis across diverse databases and domains.