Journal Article

Sentiment Analysis for Multi-Attribute Data in OSNs Using Hybrid Approach

Pallavi, P., Thallapalli, Ravikumar, Narasimha, G., Srinivas, K, Pramod Kumar, P.

2020

Overview

Increasing popularity of social networks like LinkedIn, MySpace and other networks in present days. Communication is also increased in between users present in social networks. Large amount of data being move on social media because of increase data outsourcing. Sentiment analysis is impressive and interest concept for online social networks, while different types of existing methods to find sentiment in online social networks to define communication between different users to categorize patterns with respect to similar attributes to analyze large data. We present and suggest the Hybrid Machine Learning method in this paper.(which is combination of Balanced Window and Classification based on Parts of Speech) to handle outsourced data of social networks from Face Book and other blogging services are trained and then classify the relation based on emotional aspect like positive or negative and other relations in social streams. The performance of our proposed approach is to extensively close to machine learning and identify important relevant features randomly and perform sentiment analysis in different data streams. Our experimental results show exhaustive level of classification results with comparison of existing approaches in real time environment.

• Publisher: IOP Publishing
• Subject: Classification / Data analysis / Data mining / Data transmission / Machine learning / machine learning and data pre-processing and social data streams / Online social networks / relevant data streams / Sentiment analysis / Social networks