Concept Relatedness Estimation

Concept relatedness estimation (CRE) aims to quantify the semantic similarity or association between concepts, a crucial task across diverse fields. Current research emphasizes developing robust CRE models that leverage not only pairwise relationships but also higher-order connections within graph structures, often employing graph convolutional networks or machine learning algorithms trained on both country-level and firm-level data to capture complex relationships. Improved CRE methods have significant implications for various applications, including knowledge graph embedding, information retrieval, and economic forecasting, by enabling more accurate predictions and a deeper understanding of complex systems.