SAHRAN: Sentiment Analysis of Hotel Reviews with Attention-Based Recurrent Neural Network DOI
Halit ÇETİNER, Sedat Metlek

Iğdır Üniversitesi Fen Bilimleri Enstitüsü Dergisi, Год журнала: 2025, Номер 15(1), С. 39 - 56

Опубликована: Фев. 19, 2025

Automatically analysing the sentiment of comments expressed by a user on web page for any purpose is rapidly expanding important research area. Text analysis, as it known in literature, technique that allows users to determine their emotional tendencies defined purpose. Users comment content pages used thousands people such vacation sites, shopping pages, social media, brand reviews, financial health political pages. The made have ability directly affect who wants benefit from these services way. For reasons, examine people's emotions automatic review comments. Recurrent Neural Network (RNN) based architectures achieved remarkable success solving Natural Language Processing (NLP) problems. In this article, an RNN deep learning model proposed works publicly available dataset obtained TripAdvisor and performs analysis. SAHRAN uses attention mechanism dot product structure capture words model, Bidirectional Gated Unit (BiGRU) Long Short Term Memory (BiLSTM) layers are integrated into features. As result experimental studies, performance values 0.9524, 0.9685, 0.9082 0.9338 terms precision, recall, F1 score accuracy measures, respectively.

Язык: Английский

SAHRAN: Sentiment Analysis of Hotel Reviews with Attention-Based Recurrent Neural Network DOI
Halit ÇETİNER, Sedat Metlek

Iğdır Üniversitesi Fen Bilimleri Enstitüsü Dergisi, Год журнала: 2025, Номер 15(1), С. 39 - 56

Опубликована: Фев. 19, 2025

Automatically analysing the sentiment of comments expressed by a user on web page for any purpose is rapidly expanding important research area. Text analysis, as it known in literature, technique that allows users to determine their emotional tendencies defined purpose. Users comment content pages used thousands people such vacation sites, shopping pages, social media, brand reviews, financial health political pages. The made have ability directly affect who wants benefit from these services way. For reasons, examine people's emotions automatic review comments. Recurrent Neural Network (RNN) based architectures achieved remarkable success solving Natural Language Processing (NLP) problems. In this article, an RNN deep learning model proposed works publicly available dataset obtained TripAdvisor and performs analysis. SAHRAN uses attention mechanism dot product structure capture words model, Bidirectional Gated Unit (BiGRU) Long Short Term Memory (BiLSTM) layers are integrated into features. As result experimental studies, performance values 0.9524, 0.9685, 0.9082 0.9338 terms precision, recall, F1 score accuracy measures, respectively.

Язык: Английский

Процитировано

0