DoctorateOpen Access

Algorithmic signals, sentiment space, and societal well-being through the ai lens: A case study of Türkiye

Is this your thesis?

This record came from a bulk archive import. If it’s yours, link it to your profile.

2025
0 views
0 downloads

Abstract (EN)

This thesis investigates financial volatility and spatial inequality in Türkiye through sentiment analysis, machine learning, and geospatial network analytics to assess societal quality of life and wellbeing. The study analyzes three domains using diverse datasets and methodologies. For financial analysis, daily Turkish Lira-USD exchange rates and stock market data from January 2015 to February 2024 are examined alongside sentiment analysis of international media (The Economist, The New York Times, The Guardian) and domestic sources (Yeni Şafak, Turkish Twitter). Natural language processing techniques and supervised machine learning models are employed for prediction, and XAI techniques are used for explainability. For spatial analysis, OpenStreetMap data and municipal records are used to construct a composite livability index across Istanbul's 38 districts, measuring walking accessibility in five urban dimensions: transport, green spaces, public services, safety infrastructure, and economic activity within the 15-minute city framework. The results indicate that international media sentiment demonstrates stronger associations with Turkish Lira volatility than domestic news sources. Similarly, for stock market predictions, international news sentiment consistently improves predictability across models, while local sentiment plays a lesser role. Ensemble vii learning methods achieve superior predictive accuracy compared to individual tree-based techniques. Regarding urban accessibility, while many Istanbul residents can access services within 15 minutes, access to transport and recreational facilities remains uneven. Central districts (Fatih, Beyoğlu, Kadıköy) exhibit the highest accessibility levels, while peripheral areas lag significantly. The research integrates financial sentiment analysis that captures economic stress and stability with urban sentiment analysis that measures residents' emotional and physical wellbeing, providing a holistic framework for understanding how financial volatility and spatial inequality collectively impact societal quality of life in Türkiye.

Author

Mahat Maalım Ibrahım

How to Cite

Mahat Maalım Ibrahım (Doctorate thesis). Algorithmic signals, sentiment space, and societal well-being through the ai lens: A case study of Türkiye, 2025, Ibn Haldun University.

Keywords

License

Tüm Hakları Saklıdır

This work is shared under the specified license terms.

More theses from Ibn Haldun University