The comparison of the accuracy and F1 score of TextBlob, VADER, AFINN, and the proposed approach: TriLex. We test on Sentiment140 and Amazon Review datasets corresponding to the three labels (positive, neutral, and negative).

<p>The comparison of the accuracy and F1 score of TextBlob, VADER, AFINN, and the proposed approach: TriLex. We test on Sentiment140 and Amazon Review datasets corresponding to the three labels (positive, neutral, and negative).</p>

Saved in:

Bibliographic Details
Main Author:	Abdulrahman Alharbi (21144277) (author)
Other Authors:	Rafaa Aljurbua (15237889) (author), Shelly Gupta (10527488) (author), Zoran Obradovic (834966) (author)
Published:	2025
Subjects:	Genetics Mental Health Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified short texts poses online customer reviews apply weighted averaging 2 %&# 8211 novel unsupervised approach generally short nature dynamic threshold derived improve sentiment labeling weak label based social media platforms paper proposes trilex unsupervised sentiment analysis dynamic nature sentiment analysis weak labels new label fusion approach sentiment prediction sentiment expression xlink "> wide range user sentiments textual content strong labels results demonstrate recent years multiple lexicon majority votes majority vote lstm models logistic regression labeled data f1 score existing fusion efficiency issue daily lives context limitations challenge due benchmark datasets 8 %.
Tags:	Add Tag No Tags, Be the first to tag this record!

Be the first to leave a comment!

The comparison of the accuracy and F1 score of TextBlob, VADER, AFINN, and the proposed approach: TriLex. We test on Sentiment140 and Amazon Review datasets corresponding to the three labels (positive, neutral, and negative).

Similar Items