Evaluating Public Sentiments about Turkish Universities through Twitter Using Deep Learning Techniques

Evaluating public attitudes towards Turkish institutions is critical for successful strategic decision-making and reputation management in today's dynamic educational settings. To tackle this problem, this study examines sentiment analysis of Twitter data to assess public impressions of Turkish institutions using a variety of deep learning algorithms, including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), and a hybrid model. Using a large dataset from Twitter, our technique focuses on interpreting the attitudes expressed about Turkish universities. After pre-processing the data, the suggested models are rigorously trained and tested, achieving an impressive 87.25% accuracy rate. The results give insights into popular attitudes and provide a comparative study of the models used. The findings illustrate the efficiency of deep learning approaches in identifying attitudes about Turkish institutions on Twitter, providing significant insights for handling reputation and decision-making in the educational field.

Journal

32ND IEEE SIGNAL PROCESSING AND COMMUNICATIONS APPLICATIONS CONFERENCE, SIU 2024

Year

2024

ISSN

2165-0608

DOI

10.1109/SIU61531.2024.10600970

WoS UID

WOS:001297894700199