Open AccessA novel k-nearest neighbor distance based under sampling for improved opinion mining on skewed data using random forest
Author(s) -
P Ratna Babu,
Bhanu Prakash Battula
Publication year - 2018
Publication title -
international journal of engineering and technology
Language(s) - English
Resource type - Journals
ISSN - 2227-524X
DOI - 10.14419/ijet.v7i1.8.9970
Subject(s) - computer science , class (philosophy) , sentiment analysis , data mining , k nearest neighbors algorithm , random forest , set (abstract data type) , sampling (signal processing) , social media , frame (networking) , data science , point (geometry) , artificial intelligence , machine learning , world wide web , mathematics , telecommunications , geometry , filter (signal processing) , computer vision , programming language
In recent years, consumers are performing a pilot investigation using online resources before making any decision of purchase. One of the most popular social blogging online medium is twitter. The opinions collected from twitter at any point of frame in real world scenario are tending towards class imbalance in nature. The existing algorithms for opinion mining can work better on class balance nature, where opinions (positive and negative) are almost balance. In this paper, we propose a novel approach known as Improved Opinion Mining using Under Sampling (IOMUS) to efficiently summarize the reviews of class imbalance opinion mining corpus. The experimental set up is performed on the collection of opinion mining class imbalance dataset consisting of “1155” instances. The experimental results suggest that improved performance is obtained by the proposed IOMUS algorithm than the traditional approach.