Aspect-Based Sentiment Analysis for Indonesian Tourist Attraction Reviews Using Bidirectional Long Short-Term Memory

The tourism sector in Indonesia experienced growth and made a positive contribution to the national economy, but this has yet reach its target. Therefore, government of implemented sustainable development program by establishing ten priority destinations. Aspect-based sentiment analysis (ABSA) towards tourist attraction reviews can assist developing potential goals. ABSA process compares with two deep learning models (LSTM Bi-LSTM), which are considered obtain good performance text analysis. shortcomings previous research should have examined aspect classification sequentially. This makes obtained from task invalid. Thus, study is conducted determine version model individually simultaneously. aims develop an aspect-based as intelligent system solution for applying binary relevance mechanism best LSTM or Bi-LSTM. test results showed that Bi-LSTM was superior Likewise, sequentially outperformed LSTM. average accuracy f1 score 92.22% 71,06%. Meanwhile, 90,63% precision 70,4% score.