Sentiment analysis from email pattern using feature selection algorithm

Today number of applications are available on mobile devices and computers for electronic mail (email) conversations. The demand for email communication is increasing day‐by‐day. Therefore the incoming and outgoing messages are also getting increased. However, extracting the sentiments from the emai...

Full description

Saved in:
Bibliographic Details
Published in:Expert systems Vol. 41; no. 2
Main Authors: Srinivasarao, Ulligaddala, Sharaff, Aakanksha
Format: Journal Article
Language:English
Published: Oxford Blackwell Publishing Ltd 01.02.2024
Subjects:
Algorithms
Clustering
Data mining
Electronic mail
email pattern
Feature selection
feature selection unsupervised fuzzy Gaussian mixture model
Mobile computing
Parametric analysis
Pattern analysis
Pattern classification
Sentiment analysis
sentiment clustering
Statistical analysis
Support vector machines
support vector regression
ISSN:0266-4720, 1468-0394
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Today number of applications are available on mobile devices and computers for electronic mail (email) conversations. The demand for email communication is increasing day‐by‐day. Therefore the incoming and outgoing messages are also getting increased. However, extracting the sentiments from the emails is now demanding. Therefore in the proposed method, the pattern classification and sentiment clustering are carried out in two phases. Initially, the pattern classification is performed using support vector regression, then the sentiments from such classified patterns are clustered using a unsupervised fuzzy‐model‐based Gaussian clustering algorithm. Finally, the experimental analysis is performed in Python tool. The proposed sentiment clustering from email patterns has attained a better accuracy result of 97.13%, which is found higher than other existing techniques. Along with the parametric analysis, non‐parametric statistical analysis using the Wilcoxon rank‐sum test is also carried out to identify the proposed sentiment analysis architecture's effectiveness.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0266-4720
1468-0394
DOI:10.1111/exsy.12867