The Impact of News-Based Sentiment Analysis on Bitcoin Price Prediction: A Comparison with Machine Learning Approaches

Abstract

Cryptocurrency markets differ from traditional financial instruments due to their high volatility and unpredictable nature, making price prediction studies an important area of academic research. Bitcoin is the most important cryptocurrency in terms of market value and trading volume, and it is accepted that its price movements are driven not only by technical indicators but also by investor sentiment and behavioral factors. In this context, investigating the role of sentiment analysis-based indicators in price prediction processes could contribute to both literature and practical applications. In this study, the impact of sentiment analysis derived from news content on Bitcoin price prediction was examined. Four different machine learning algorithms—Linear Regression, Rastgele Orman, Support Vector Regression, and XGBoost Regressor—were used, and the models were tested under two different scenarios created with technical data and the integration of sentiment scores. The findings revealed that ensemble-based methods, in particular, benefited significantly from sentimental data. Before improvement, the Support Vector Regression model produced the lowest error values with 5.024 MAE and 7.257 RMSE, while the XGBoost model showed weaker performance with 6.109 MAE and 7.986 RMSE. After integrating sentiment analysis, the error values decreased to 490 MAE and 772 RMSE in the Rastgele Orman model and 676 MAE and 1,137 RMSE in the XGBoost model. In contrast, the Linear Regression model showed an overfitting tendency with 0.00 error values, while Support Vector Regression provided limited improvement with 4.810 MAE and 6.960 RMSE. Overall, the findings suggest that considering news-based sentiment signals in price prediction models can increase predictive power. In this regard, the study contributes to a more comprehensive understanding of cryptocurrency markets through a behavioral finance approach.

Keywords

• Cryptocurrency
• Natural language processing
• Sentiment analysis
• Machine learning
• Time series.

Haber Tabanlı Duygu Analizinin Bitcoin Fiyat Tahminine Etkisi: Makine Öğrenmesi Yaklaşımlarıyla Bir Karşılaştırma

Abstract

Kripto para piyasaları, yüksek volatilite ve öngörülemez yapıları nedeniyle geleneksel finansal araçlardan ayrışmakta ve bu durum fiyat tahmini çalışmalarını akademik açıdan önemli bir araştırma alanı haline getirmektedir. Bitcoin, piyasa değeri ve işlem hacmi bakımından öne çıkan en önemli kripto varlık olup, fiyat hareketlerinin yalnızca teknik göstergelerle değil, aynı zamanda yatırımcı duyarlılığı ve davranışsal faktörlerle de yönlendirildiği kabul edilmektedir. Bu bağlamda, fiyat tahmini süreçlerinde duygu analizi tabanlı göstergelerin rolünü araştırmak hem literatüre hem de pratik uygulamalara katkı sunabilecek niteliktedir. Bu çalışmada, haber içeriklerinden elde edilen duygu analizlerinin Bitcoin fiyat tahminine etkisi incelenmiştir. Lineer Regresyon, Rastgele Orman, Destek Vektör Regresyonu ve XGBoost Regressor olmak üzere dört farklı makine öğrenimi algoritması kullanılmış, modeller teknik verilerle ve duygu skorlarının entegrasyonu ile oluşturulan iki farklı senaryo altında test edilmiştir. Bulgular, özellikle topluluk yöntemleri (ensemble) tabanlı yöntemlerin duygu verilerinden anlamlı ölçüde faydalandığını ortaya koymuştur. İyileştirme öncesinde Destek Vektör Regresyon modeli 5.024 MAE ve 7.257 RMSE ile en düşük hata değerini üretirken, XGBoost modeli 6.109 MAE ve 7.986 RMSE ile daha zayıf performans sergilemiştir. Duygu analizi entegrasyonu sonrasında ise Rastgele Orman modelinde hata değerleri 490 MAE ve 772 RMSE’ye, XGBoost modelinde ise 676 MAE ve 1.137 RMSE’ye kadar düşmüştür. Buna karşılık, Lineer Regresyon modeli 0.00 hata değerleriyle aşırı öğrenme eğilimi göstermiş, Destek Vektör Regresyon ise 4.810 MAE ve 6.960 RMSE ile sınırlı iyileşme sağlamıştır.

Keywords

• Kripto para
• Doğal dil işleme
• Duygu analizi
• Makine öğrenmesi
• Zaman serisi.

References

1. Arlot, S., & Celisse, A. (2010). A survey of cross-validation procedures for model selection. Statistics Surveys, 4, 40–79. https://doi.org/10.1214/09-SS054
2. Bollen, J., Mao, H., & Zeng, X. (2011). Twitter mood predicts the stock market. Journal of Computational Science, 2(1), 1–8. https://doi.org/10.1016/j.jocs.2010.12.007 Buathong, W., Sieng-EK, P., & Jarupunphol, P. (2023). Measuring the performance of machine learning forecasting models to support bitcoin investment decisions [J]. Journal of Data Science and Intelligent Systems.
3. Catalini, C., & Gans, J. S. (2016). Some simple economics of the blockchain. National Bureau of Economic Research (NBER) Working Paper No. 22952. https://doi.org/10.3386/w22952 Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785–794. https://doi.org/10.1145/2939672.2939785 Corbet, S., Lucey, B., Urquhart, A., & Yarovaya, L. (2019). Cryptocurrencies as a financial asset: A systematic analysis. International Review of Financial Analysis, 62, 182–199. https://doi.org/10.1016/j.irfa.2018.09.003 Dyhrberg, A. H. (2016). Bitcoin, gold and the dollar – A GARCH volatility analysis. Finance Research Letters, 16, 85–92. https://doi.org/10.1016/j.frl.2015.10.008
4. Garcia, D., Tessone, C. J., Mavrodiev, P., & Perony, N. (2014). The digital traces of bubbles: Feedback cycles between socio-economic signals in the Bitcoin economy. Journal of the Royal Society Interface, 11(99), 20140623. https://doi.org/10.1098/rsif.2014.0623
5. Hutto, C., & Gilbert, E. (2014). Vader: A parsimonious rule-based model for sentiment analysis of social media text. In Proceedings of the international AAAI conference on web and social media (Vol. 8, No. 1, pp. 216-225).
6. Hyndman, R. J., & Athanasopoulos, G. (2018). Forecasting: principles and practice. OTexts.
7. Kaggle. (2025a). Cryptonews 2019–2023 [Veri seti]. Kaggle. Erişim tarihi: 10 Ağustos 2025, https://www.kaggle.com/datasets/larysa21/cryptonews-2019-2023
8. Kaggle. (2025b). Crypto currencies daily prices [Veri seti]. Kaggle. Erişim tarihi: 10 Ağustos 2025, https://www.kaggle.com/datasets/svaningelgem/crypto-currencies-daily-prices?select=BTC.csv

9. Kaminski, J. (2014). Nowcasting the Bitcoin market with Twitter signals. arXiv preprint, arXiv:1406.7577.
10. Kim, Y., Kang, S., & Kim, S. (2016). Financial news analysis using machine learning techniques: An ensemble approach. Expert Systems with Applications, 62, 205–215.
11. Kim, K. (2020). Cryptocurrency trader's risky decision on social sentiment. In GITMA 2020: 19th Global Information Technology Management Association (Virtual Conference).
12. Kohavi, R. (1995). A study of cross-validation and bootstrap for accuracy estimation and model selection. In Ijcai (Vol. 14, No. 2, pp. 1137-1145).
13. Kristoufek, L. (2013). Bitcoin meets Google Trends and Wikipedia: Quantifying the relationship between phenomena of the Internet era. Scientific Reports, 3(1), 3415. https://doi.org/10.1038/srep03415
14. Liu, B. (2012). Sentiment analysis and opinion mining. Synthesis Lectures on Human Language Technologies, 5(1), 1–167. https://doi.org/10.2200/S00416ED1V01Y201204HLT016
15. Mai, F., Shan, Z., Bai, Q., Wang, X., & Chiang, R. H. L. (2018). How does social media impact Bitcoin value? A test of the silent majority hypothesis. Journal of Management Information Systems, 35(1), 19–52. https://doi.org/10.1080/07421222.2018.1440774
16. Mao, H., Counts, S., & Bollen, J. (2015). Quantifying the effects of online bullishness on international financial markets (No. 9). ECB Statistics Paper.
17. Montgomery, D. C., Peck, E. A., & Vining, G. G. (2012). Introduction to linear regression analysis (5th ed.). John Wiley & Sons.
18. Nassirtoussi, A. K., Aghabozorgi, S., Wah, T. Y., & Ngo, D. C. L. (2014). Text mining for market prediction: A systematic review. Expert Systems with Applications, 41(16), 7653–7670. https://doi.org/10.1016/j.eswa.2014.06.009
19. Phillips, R. C., & Gorse, D. (2017). Predicting cryptocurrency price bubbles using social media data and epidemic modelling. In 2017 IEEE symposium series on computational intelligence (SSCI) (pp. 1-7). IEEE. https://doi.org/10.1109/SSCI.2017.8280809
20. Şahinaslan, Ö., Dalyan, H., & Şahinaslan, E. (2022). Naive bayes sınıflandırıcısı kullanılarak youtube verileri üzerinden çok dilli duygu analizi. Bilişim Teknolojileri Dergisi, 15(2), 221-229.
21. Smola, A. J., & Schölkopf, B. (2004). A tutorial on support vector regression. Statistics and Computing, 14(3), 199–222. https://doi.org/10.1023/B:STCO.0000035301.49549.88
22. Yermack, D. (2024). Is Bitcoin a real currency? An economic appraisal. In Handbook of digital currency (pp. 29-40). Academic Press.