Decoding Weibo Sentiments: Unveiling Nuanced Emotions with Bidirectional LSTM Analysis

Kaiwen Deng

2024

Abstract

This study highlights the critical role that deep learning plays in deciphering the complex emotional subtleties prevalent in social media interactions via sentiment analysis of Weibo comments. The primary goal is to deploy a bidirectional Long Short-Term Memory (LSTM) model that is intended to thoroughly analyze and understand user attitudes, providing priceless information for improving marketing tactics and gauging public opinion. Utilizing character-level segmentation and Tencent Artificial Intelligence (AI) Lab's pre-trained word embeddings, the study advances sentiment analysis by enhancing sensitivity to contextual subtleties within textual data. The study, which used a dataset of Weibo comments with sentiment annotations, demonstrates the remarkable 96% accuracy with which the bidirectional LSTM model can categorize sentiments. This result demonstrates how well the model captures complex emotional expressions, outperforming both other deep learning methods and conventional machine learning techniques in sentiment analysis tasks. The novel elements of the model's architecture, such character-level analysis and the intelligent use of pre-trained embeddings, improve its classification accuracy and contextual comprehension. These aspects represent significant advancements in sentiment analysis, with broad implications for both academic research and practical applications in understanding social media discourse.

Paper Citation

in Harvard Style

Deng K. (2024). Decoding Weibo Sentiments: Unveiling Nuanced Emotions with Bidirectional LSTM Analysis. In Proceedings of the 1st International Conference on Engineering Management, Information Technology and Intelligence - Volume 1: EMITI; ISBN 978-989-758-713-9, SciTePress, pages 221-226. DOI: 10.5220/0012923300004508

in Bibtex Style

@conference{emiti24,
author={Kaiwen Deng},
title={Decoding Weibo Sentiments: Unveiling Nuanced Emotions with Bidirectional LSTM Analysis},
booktitle={Proceedings of the 1st International Conference on Engineering Management, Information Technology and Intelligence - Volume 1: EMITI},
year={2024},
pages={221-226},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012923300004508},
isbn={978-989-758-713-9},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Engineering Management, Information Technology and Intelligence - Volume 1: EMITI
TI - Decoding Weibo Sentiments: Unveiling Nuanced Emotions with Bidirectional LSTM Analysis
SN - 978-989-758-713-9
AU - Deng K.
PY - 2024
SP - 221
EP - 226
DO - 10.5220/0012923300004508
PB - SciTePress