Review Paper: Social Media Sentiment Analysis Using Twitter Dataset

Review Paper: Social Media Sentiment Analysis Using Twitter Dataset

Swati Kashyap, Sushant Kumar, Vishal Vyas, Tushar Anand, Raman Kumar

BE-CSE
Chandigarh University
Mohali, Punjab

Abstract— The expansion of social media platforms, particularly Twitter, has transformed them into vast databases of user-generated content expressing a wide range of thoughts and emotions. This review delves into the field of conceptual analysis applied to Twitter datasets, focusing on machine learning. This article provides an overview of methods, techniques, and operational models for sentiment analysis using machine learning in the specific context of Twitter.

Introducing the evolution of machine learning applied to the Twitter dataset, starting by exploring the main concepts of sentiment analysis. A variety of supervised and unsupervised learning methods are reviewed, from traditional methods to recent developments including deep learning models. This article discusses the key features and issues associated with this process, considering abbreviations, nuances of user-generated messages, and quality of Twitter content.

This research includes a discussion of structure, lexical theory, and the role of pre-learning word embeddings in improving the performance of learning models. Focus on transforming the structure into specific ideas and connecting the data of the points to achieve higher classification.

Practical uses of sentiment analysis, including but not limited to understanding public sentiment on Twitter, tracking product sentiment, and finding trends, are a good fit. The review also addresses ethical considerations and issues related to bias in the perspective of analyzing patterns studied in Twitter data.

The article concludes by presenting the current state, open challenges and future directions in the Twitter space using machine learning. Combining insights from existing literature, this review provides useful material for researchers, practitioners, and enthusiasts exploring the complexities of adapting to the Twitter environment and analyzing sentiment in power.

The widespread reach of social media platforms like Twitter gives them the advantage of rich, original user-generated content that reflects different perspectives and opinions. This in-depth review article aims to explore the dynamic areas of sentiment analysis applied to Twitter datasets, with particular emphasis on the application of different types of machine learning. The narrative turns the complexity of sentiment analysis on its head by delving into the nuances of Twitter's dynamic content and the evolution of machine learning.

This article begins with a detailed review of the fundamentals of sentiment analysis and examines the evolution of machine learning techniques adapted to the Twitter dataset. Panoramic research covers many methods, from classical supervised and unsupervised models to deep learning models. The article will take a closer look at the specific challenges encountered in analysing short texts, user-generated messages, and modifying messages on Twitter.