Title AI-based monitoring of fdi perceptions: developing a real-time reputation index for investment promotion agencies in the baltic states
Translation of Title Operatyvus informacijos vertinimas naudojant DI: TUI reputacijos indekso matavimo sistema Baltijos šalims.
Authors Balodis, Marius
Full Text Download
Pages 115
Keywords [eng] foreign direct investment, text embeddings, google search volume, media monitoring, narrative analysis, nation's image, reputation management fdi, artificial intelligence, dirbtinis intelektas, tui, tiesioginės užsienio investicijos, žiniasklaidos monitoringas, šalies įvaizdis, reputacijos valdymas.
Abstract [eng] This thesis investigates the application of real-time media exposure and sentiment signals to assess foreign direct investment (FDI) attractiveness for investment promotion activities in the Baltic States. The work focuses on the growing mismatch between rapidly changing perception dynamics and traditional FDI evaluation tools, which rely on lagged statistical indicators and infrequent surveys. The scientific problem is the absence of a platform capable of continuously monitoring FDI-related perceptions of volatile entities in mass media. The thesis objective is to develop and test a real-time investment reputation index based on AI-driven media analytics to support investment promotion agencies (IPAs). To achieve this objective, the study first analyses the determinants of FDI and investor perceptions through a literature review, then identifies FDI-relevant communication narratives and examines them across passive and active media exposure components provided by a professional media monitoring provider. Finally, the study employs natural language processing methods, as well as sentiment, narrative, prominence, and quantitative interest analysis to construct a country-level FDI perception index to examine investment image patterns across the three Baltic States. The findings indicate that active perception indicators, including search-driven visibility and curated information environments, are more closely aligned with investment-related attention than passive media exposure alone. Comparative analysis reveals distinct FDI-related reputation profiles across the Baltic States, with Lithuania demonstrating the strongest overall performance in the constructed index and independent investment assessments by popular large language models (LLMs). The conclusions confirm that real-time perception monitoring works alongside traditional FDI indicators and provides practical value as an early-warning and reputation-management tool for IPAs. Based on the findings, practical recommendations are formulated to improve investment communication, narrative targeting, and strategic reputation management. The results are intended for application in investment promotion practice and further development of AI-based monitoring solutions.
Dissertation Institution Vilniaus universitetas.
Type Master thesis
Language English
Publication date 2026