Skripsi
ANALISIS SENTIMEN PADA ULASAN PENGGUNA APLIKASI AJAIB DENGAN ALGORITMA LONG SHORT-TERM MEMORY
Ajaib is one of the largest applications in Indonesia that focuses on stock and mutual fund investments. User reviews of the application are one way for Ajaib to understand user needs and make improvements to the application. This research aims to analyze the sentiment in user reviews of the Ajaib application. Long Short-Term Memory (LSTM) is used as the method for sentiment analysis, and Word2Vec is used for feature extraction. The data used in this research consists of 1,111 user reviews of the Ajaib application from the Google Play Store. The testing phase includes 6 experimental scenarios, and the best results were obtained with the following hyperparameters: dropout layer value is 0.2, 32 LSTM neurons, learning rate value is 10-3, LSTM dropout value is 0.5, batch size value is 8, and 30 epochs. The results show the highest accuracy and f-measure values is 89.69% and 89.72%.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2307002537 | T114317 | T1143172023 | Central Library (Referens) | Available |
No other version available