PREDIKSI TREN HARGA EUR DAN USD BERDASARKAN ANALISIS DATA BERITA FINANSIAL MENGGUNAKAN MODEL LONGFORMER-BILSTM

Muhammad Saiful Millah; Rudi Kurniawan; Bani Nurhakim; Denni Pratama

Authors

Muhammad Saiful Millah STMIK IKMI Cirebon, Indonesia
Rudi Kurniawan STMIK IKMI Cirebon, Indonesia
Bani Nurhakim STMIK IKMI Cirebon, Indonesia
Denni Pratama STMIK IKMI Cirebon, Indonesia

Keywords:

Longformer–BiLSTM, analisis sentimen finansial, prediksi tren EUR, USD

Abstract

Penelitian ini bertujuan untuk meningkatkan akurasi prediksi tren nilai tukar EUR/USD melalui pengembangan model hibrida Longformer–BiLSTM yang memanfaatkan analisis sentimen berita finansial. Pergerakan EUR/USD sangat dipengaruhi oleh kondisi makroekonomi global dan sentimen pasar, sehingga integrasi data harga historis dan sentimen berita dipandang mampu memberikan gambaran yang lebih komprehensif mengenai dinamika pasar valuta asing. Dataset yang digunakan mencakup artikel berita finansial berbahasa Inggris dari Reuters, Bloomberg, dan Investing.com yang disejajarkan secara temporal dengan data harga penutupan harian EUR/USD. Prapengolahan data meliputi pembersihan teks, tokenisasi, ekstraksi sentimen menggunakan FinBERT, normalisasi fitur, serta pembentukan jendela waktu sepanjang 60 hari untuk membangun urutan input model. Arsitektur Longformer dimanfaatkan untuk menangani dokumen panjang dan mengekstraksi representasi semantik yang mendalam, sedangkan BiLSTM digunakan untuk memodelkan dependensi temporal dari gabungan data harga dan sentimen. Model dilatih menggunakan pembagian data berurutan (training–validation–testing) serta penyetelan hiperparameter untuk memperoleh konfigurasi optimal. Evaluasi performa dilakukan menggunakan RMSE dan MAPE, kemudian dibandingkan dengan beberapa model baseline seperti LSTM, GRU, dan CNN–BiLSTM. Hasil penelitian menunjukkan bahwa Longformer–BiLSTM unggul secara signifikan dengan RMSE sekitar 0,042, MAPE 1,84%, serta nilai R² mendekati 0,982. Kinerja tersebut terbukti lebih baik dibandingkan seluruh model pembanding, menunjukkan bahwa integrasi informasi sentimen berita ke dalam arsitektur berbasis transformer dan jaringan BiLSTM mampu meningkatkan akurasi, stabilitas, dan kemampuan generalisasi prediksi tren EUR/USD. Penelitian ini memberikan kontribusi berupa kerangka multimodal yang efektif dan dapat direplikasi untuk peramalan pasar valuta asing berbasis data dan sentimen.

References

Adelakun & Baale A. A., N. O. (2024). Sentiment analysis of financial news using the BERT model. ITEGAM-JETIA, 10(48), 21-27. https://doi.org/https://doi.org/10.5935/jetia.v10i48.1029

Ahmad Reynolds J. & Rezgui Y., M. W. (2022). Evaluation of regression performance metrics for time-series prediction. Applied Energy, 305, 117932. https://doi.org/https://doi.org/10.1016/j.apenergy.2021.117932

Ahmed & Zhao Y., M. (2024). Interactive visualization of sentiment-price correlation in hybrid forecasting models. Applied Soft Computing, 152, 110189. https://doi.org/https://doi.org/10.1016/j.asoc.2024.110189

Ahmed Abdalla Mohammed B, P. (2023). A., & Saeed, A. M. (2023). The Impact of Image Augmentation Techniques of MRI Patients in Deep Transfer Learning Networks for Brain Tumour Detection. Journal of Electrical Systems and Information Technology. https://doi.org/10.1186/s43067-023-00119-9

Araci, D. (2019). FinBERT: Financial sentiment analysis with pre-trained language models. ArXiv Preprint ArXiv:1908.10063. https://doi.org/https://doi.org/10.48550/arXiv.1908.10063

Arif Herlinawati E. Devianto D. Yollanda M. & Permana D., E. (2024). Hybridization of long short-term memory neural network in fractional time series modeling of inflation. Frontiers in Big Data, 6, 1282541. https://doi.org/https://doi.org/10.3389/fdata.2023.1282541

Ayitey Junior, M., Appiahene, P., Appiah, O., & Bombie, C. N. (2023). Forex market forecasting using machine learning: Systematic Literature Review and meta-analysis. Journal of Big Data, 10(1). https://doi.org/10.1186/s40537-022-00676-2

Beltagy Peters M. E. & Cohan A., I. (2020). Longformer: The long-document transformer. ArXiv Preprint. https://doi.org/https://doi.org/10.48550/arXiv.2004.05150

Bukhari Raja M. A. Z. Sulaiman M. Islam S. Shoaib M. & Kumam P., A. H. (2020). Fractional neuro-sequential ARFIMA-LSTM for financial market forecasting. IEEE Access, 8, 71326–71338. https://doi.org/https://doi.org/10.1109/ACCESS.2020.2985763

Cao Zhang T. & Huang J., K. (2024). Advanced hybrid LSTM-Transformer architecture for real-time multi-task prediction in engineering systems. Scientific Reports, 14, 4890. https://doi.org/https://doi.org/10.1038/s41598-024-4890

Chahuán-Jiménez, K. (2024). Neural Network-Based Predictive Models for Stock Market Index Forecasting. Journal of Risk and Financial Management, 17(6). https://doi.org/10.3390/jrfm17060242

Chen Wu X. & Zhao J., L. (2024). Performance evaluation of hybrid forecasting models using multiple error indicators. Frontiers in Artificial Intelligence, 7, 1392447. https://doi.org/https://doi.org/10.3389/frai.2024.1392447

Chen Xu R. & Yang D., J. (2024). Learning rate scheduling and optimization strategies in transformer–RNN hybrid models. Neural Computing and Applications, 36, 10271–10284. https://doi.org/https://doi.org/10.1007/s00521-023-09056-5

Chen Zhao Q. & Zhang T., R. (2022). Comparative quantitative design in deep learning models for stock and forex forecasting. Information Sciences, 602, 320–338. https://doi.org/https://doi.org/10.1016/j.ins.2022.04.073

Dahal Pokhrel N. R. Gaire S. Mahatara S. Joshi R. P. Gupta A. Banjade H. R. & Joshi J., K. R. (2023). A comparative study on effect of news sentiment on stock price prediction with deep learning architecture. PLOS ONE, 18(4), E0284695. https://doi.org/https://doi.org/10.1371/journal.pone.0284695

Daradkeh, M. K. (2022). A Hybrid Data Analytics Framework with Sentiment Convergence and Multi-Feature Fusion for Stock Trend Prediction. Electronics (Switzerland), 11(2). https://doi.org/10.3390/electronics11020250

Du Xing F. Mao R. & Cambria E., K. (2024). Financial sentiment analysis: Techniques and applications. ACM Computing Surveys. https://doi.org/https://doi.org/10.1145/3649451

Fazlija & Harder P., B. (2022). Using financial news sentiment for stock price direction prediction. Mathematics, 10(13), 2156. https://doi.org/https://doi.org/10.3390/math10132156.

Filippou Taylor M. P. & Wang Z., I. (2024). Media sentiment and currency reversals. Journal of Financial and Quantitative Analysis, 59(3), 1401–1429. https://doi.org/https://doi.org/10.1017/S0022109023000534

Fisichella & Garolla F., M. (2021). Can deep learning improve technical analysis of Forex data to predict future price movements? IEEE Access. Https://Doi.Org/10.1109/ACCESS.2021.3127570. https://doi.org/https://doi.org/10.1109/ACCESS.2021.3127570

Groß-Klußmann, A. (2024). Learning deep news sentiment representations for macro-finance. Digital Finance, 6(3), 341–377. https://doi.org/10.1007/s42521-024-00107-2

Hizzard et al., S. (2023). Forex market forecasting using machine learning: Systematic literature review and meta-analysis. Journal of Big Data, 10, Article 9. https://doi.org/https://doi.org/10.1186/s40537-022-00676-2

Hodson, T. O. (2022). Root-mean-Squared error (RMSE) or mean absolute error (MAE): When to use them or not. Geoscientific Model Development, 15, 5481–5487. https://doi.org/https://doi.org/10.5194/gmd-15-5481-2022

Hu et al, Q. (2021). (2021). Improved Classification of Benign and Malignant Breast Lesions Using Deep Feature Maximum Intensity Projection MRI in Breast Cancer Diagnosis. Radiology: Artificial Intelligence. https://doi.org/10.1148/ryai.2021200159

Hu Zhao Y. & Khushi M., Z. (2021). A survey of forex and stock price prediction using deep learning. Applied System Innovation, 4(1), 9. https://doi.org/https://doi.org/10.3390/asi4010009.

Hyndman & Koehler A. B., R. J. (2020). Another look at measures of forecast accuracy. International Journal of Forecasting, 36(1), 240–245. https://doi.org/https://doi.org/10.1016/j.ijforecast.2019.05.011

Ismail, M. T. (2023). Forex market forecasting using machine learning: Systematic literature review and meta-analysis. Journal of Big Data, 10, 9. https://doi.org/https://doi.org/10.1186/s40537-022-00676-2

Li & Wang S., X. (2021). Quantitative evaluation of hybrid deep learning models for financial time-series prediction. IEEE Access, 9, 77145–77160. https://doi.org/https://doi.org/10.1109/ACCESS.2021.3089042

Li Li Y. & Wang S., X. (2020). Visualization strategies for time-series forecasting evaluation. Expert Systems with Applications, 147, 113211. https://doi.org/https://doi.org/10.1016/j.eswa.2020.113211

Li Liu Y. & Wang S, X. (2020). Visual analytics for exploring relationships between sentiment and financial time series. Expert Systems with Applications, 146, 113153. https://doi.org/https://doi.org/10.1016/j.eswa.2020.113153

Li Wang Y. & Zhao X., J. (2021). Hybrid transformer-BiLSTM networks for long text sentiment analysis. IEEE Access, 9, 139271–139283. https://doi.org/https://doi.org/10.1109/ACCESS.2021.3119134

Liu, X. (2023). (Freddie), Karagoz, G., & Meratnia, N. (2023). Analyzing the Impact of Data Augmentation on the Explainability of Deep Learning-Based Medical Image Classification. Machine Learning and Knowledge Extraction. https://doi.org/10.3390/make7010001

Liu Xu Y. & Zhao P., Z. (2021). Evaluating sentiment-based financial models through experimental research frameworks. IEEE Access, 9, 117540–117556. https://doi.org/https://doi.org/10.1109/ACCESS.2021.3093428

Makridakis Spiliotis E. & Assimakopoulos V., S. (2021). The M4 competition: Results, findings, conclusion, and way forward. International Journal of Forecasting, 37(2), 789–808. https://doi.org/https://doi.org/10.1016/j.ijforecast.2020.09.006

Memiş, E., Akarkamçı, H., Yeniad, M., Rahebi, J., & Lopez-Guede, J. M. (2024). Comparative Study for Sentiment Analysis of Financial Tweets with Deep Learning Methods. Applied Sciences (Switzerland), 14(2). https://doi.org/10.3390/app14020588

Nasiopoulos Roumeliotis K. I. Sakas D. P. Toudas K. & Reklitis P., D. K. (2025). Financial sentiment analysis and classification: A comparative study of fine-tuned deep learning models. International Journal of Financial Studies, 13(2), 75. https://doi.org/https://doi.org/10.3390/ijfs13020075

Palomino & Aider F., M. A. (2022). Evaluating the effectiveness of text pre-processing in sentiment analysis. Applied Sciences, 12(17), 8765. https://doi.org/https://doi.org/10.3390/app12178765

Pang Ju W. Welch M. Gauger P. Liu P. Zhang Q. & Wang C, J. (2023). (2023). Nonparametric Bootstrap Methods for Interval Estimation of the Area under the ROC Curve with Correlated Diagnostic Test Data: Application to Whole-Virus ELISA Testing in Swine. Frontiers in Veterinary Science. https://doi.org/10.3389/fvets.2023.1274786

Parekh Patel N. P. Thakkar N. Gupta R. Tanwar S. Sharma G. Davidson I. E. & Sharma R., R. (2022). DL-GuesS: Deep learning and sentiment analysis-based cryptocurrency price prediction. IEEE Access, 10, 35398–35409. https://doi.org/https://doi.org/10.1109/ACCESS.2022.3163305

Siino, M. (2024). Is text preprocessing still worth the time? A comparative survey on text preprocessing for classification tasks. Pattern Recognition and Artificial Intelligence, 80, 109538. https://doi.org/https://doi.org/10.1016/j.patcog.2023.109538

Singh Kaur H. & Li D., P. (2023). Dynamic correlation visualization for financial sentiment analysis. Frontiers in Artificial Intelligence, 6, 1162458. https://doi.org/https://doi.org/10.3389/frai.2023.1162458

Singh, R., & al., et. (2022). Overcoming data imbalance in medical image classification using augmentation techniques. Artificial Intelligence in Medicine.

Smith, L. N. (2020). Cyclical learning rates for training neural networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(5), 1859–1871. https://doi.org/https://doi.org/10.1109/TPAMI.2020.2999397

Song, S. (2023). Deep learning-driven text sentiment analysis: Research progress, challenges, and future trends in the past five years. Frontiers in Computing and Intelligent Systems, 2, Article 31919. https://doi.org/https://doi.org/10.54097/9y4deh13

Sun Wei Y. & Lai K. K., S. (2022). Exchange rate forecasting with online forex news sentiment mining. China Journal of Econometrics, 2(2), 441-464. https://doi.org/https://doi.org/10.12012/CJoE2022-0016

Vickers, A. (2021). J. (2021). Pitfalls in Training and Validation of Deep Learning Systems. Endoscopy. https://doi.org/10.1055/a-1414-3529

Wang & Xu L., Y. (2022). Integrating economic data sources for forex trend forecasting. Information Sciences, 607, 317–333. https://doi.org/https://doi.org/10.1016/j.ins.2022.06.109

Wang Chen T. & Zhao Q., R. (2022). Correlation visualization for sentiment-based stock and forex forecasting. Information Sciences, 601, 700–715. https://doi.org/https://doi.org/10.1016/j.ins.2022.03.090

Wang Wang X. Li J. & Wang H., J. (2021). A prediction model of CNN-TLSTM for USD/CNY exchange rate prediction. IEEE Access, 9, 73346–73354. https://doi.org/https://doi.org/10.1109/ACCESS.2021.3080459

Xiong Chen G. & Cao J., Y. (2024). Research on public service request text classification based on BERT-BiLSTM-CNN feature fusion. Applied Sciences, 14(14), 6282. https://doi.org/https://doi.org/10.3390/app14146282

Xu, W. (2023). (2023). ResNet and Its Application to Medical Image Processing. Pattern Recognition Letters. https://doi.org/10.1016/j.patrec.2023.03.025

Yang et al., M. (2022). News text-mining-based business sentiment analysis and its significance in economy. Sustainability, 14(18), 11234. https://doi.org/https://doi.org/10.3390/su141811234

Yu, C., & Pei, H. (2024). Dynamic Weighting Translation Transfer Learning for Imbalanced Medical Image Classification. Entropy, 26(5). https://doi.org/10.3390/e26050400

Zhai Li J. & Wong C., X. (2022). Domain-adaptive BERT for financial sentiment analysis. Information Processing & Management, 59(5), 103037. https://doi.org/https://doi.org/10.1016/j.ipm.2022.103037

Zhang & Huang D., W. (2023). Experimental validation of hybrid deep learning models for financial sentiment forecasting. Frontiers in Artificial Intelligence, 6, 1153412. https://doi.org/https://doi.org/10.3389/frai.2023.1153412

Zhang & Xu L., J. (2021). Sentiment–market co-movement visualization using time-aligned correlation analysis. IEEE Access, 9, 88310–88324. https://doi.org/https://doi.org/10.1109/ACCESS.2021.3092208

Zhang et al., L. (2025). Major issues in high-frequency financial data analysis. Mathematics, 13(3), 347. https://doi.org/https://doi.org/10.3390/math13030347

Zhang, J., & al., et. (2020). Domain adaptation in medical imaging: A systematic review. IEEE Transactions on Medical Imaging.

Zhou Chen Q. & Wang L., T. (2023). Synchronizing multi-source financial data streams for predictive modeling. Frontiers in Artificial Intelligence, 6, 1145987. https://doi.org/https://doi.org/10.3389/frai.2023.1145987