Home // IMMM 2017, The Seventh International Conference on Advances in Information Mining and Management // View article

Improving Twitter Sentiment Classification Using Term Usage And User Based Attributes

Authors:
Selim Akyokuş
Murat Can Ganiz
Cem Gümüş

Keywords: Sentiment Analysis; Sentiment Classification; Machine Learning; Feature Engineering; Feature Extraction;

Abstract:
With the rapid growth of the Internet and the increase in the use of mobile devices, social media has grown rapidly in recent years. Without using appropriate representation techniques, processing methods and algorithms, it is difficult to get patterns, trends and opinions that are of interest to companies, organizations and individuals. Sentiment classification, which is one of the most popular mining tasks on the textual part of the social media data, aims to classify comment texts by their polarity. Textual features such as terms, n-grams combined with the NLP techniques are commonly used for this task. Our aim in this study is to see the effect of additional features on Twitter sentiment classification that are extracted from structured data related to the tweets and the Twitter users associated with these tweets. In addition to the use of terms in tweets as features i.e. traditional bag-of-words model, we employed tweet term usage based attributes along with Twitter user based attributes and showed that these additional attributes increase the accuracy of class substantially.

Pages: 47 to 51

Copyright: Copyright (c) IARIA, 2017

Publication date: June 25, 2017

Published in: conference

ISSN: 2326-9332

ISBN: 978-1-61208-566-1

Location: Venice, Italy

Dates: from June 25, 2017 to June 29, 2017