Economy and Geography Shape the Collective Attention of Cities

TL;DR

This study investigates how economic and geographic factors influence collective attention in cities, revealing that wealthier cities tend to be central and influential within their attention networks, with regional differences in population effects.

Contribution

It introduces a comprehensive analysis combining attention networks with economic and geographic data, highlighting the roles of GDP and location in urban attention patterns.

Findings

High-GDP cities act as core nodes in attention networks.

Geographic proximity and topical coherence are linked in urban clusters.

Population size influences attention in the US but not in China.

Abstract

Complex networks are commonly used to explore human behavior. However, previous studies largely overlooked the geographical and economic factors embedded in collective attention. To address this, we construct attention networks from time-series data for the United States and China, each a key economic power in the West and the East, respectively. We reveal a strong macroscale correlation between urban attention and Gross Domestic Product (GDP). At the mesoscale, community detection of attention networks shows that high-GDP cities consistently act as core nodes within their communities and occupy strategic geographic positions. At the microscale, structural hole theory identifies these cities as key connectors between communities, with influence proportional to economic output. Overlapping community detection further reveals tightly connected urban clusters, prompting us to introduce geographic and topic-based metrics, which show that closely linked cities are spatially proximate and topically coherent. Of course, not all patterns were consistent across regions. A notable distinction emerged in the relationship between population size and urban attention, which was evident in the United States but absent in China. Building on these insights, we integrate key variables reflecting GDP, geography, and scenic resources into regression model to cross-verify the influence of economic and geographic factors on collective user attention, and unexpectedly discover that a composite index of population, access, and scenery fails to account for cross-city variations in attention. Our study bridges the gap between economic prosperity and geographic centrality in shaping urban attention landscapes.