Scholarlink

<p>Numerous studies have evaluated the quality of Volunteered Geographic Information (VGI) using both extrinsic and intrinsic methods. This paper presents a new approach that combines both evaluations. The method introduces a feature matching technique that uses fuzzy normalized values of three geometric criteria: Hausdorff Distance, Direction Difference, and Area of Buffers’ Overlap. The study analyzes OpenStreetMap (OSM) history data for the study area and assesses the positional accuracy of the initial and most recent versions of OSM features by comparing them with an official reference dataset. As a primary objective, contribution volume, the number of individual contributors, and population density are examined as potential indicators of data quality. The analysis of contribution volume is carried out in four steps. First, comparing the initial and latest versions of features shows a significant improvement in positional accuracy, suggesting that repeated edits help refine data over time. Second, when features are grouped by usage type, accuracy improvements are observed, although the degree of change differs across categories. Third, analyzing features by their number of versions reveals a positive relationship between revision count and accuracy growth. Fourth, results from the latest dataset indicate a threshold effect, where additional revisions do not always lead to further improvement. In contrast, contributor count does not demonstrate a statistically significant association with positional accuracy, challenging the assumption that a higher number of contributors necessarily improves quality. Similarly, features in lower-density areas show slightly higher accuracy, suggesting that contextual factors may outweigh population concentration effects.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A novel framework for evaluating the influence of participatory contributions and demographic factors on the positional accuracy of OpenStreetMap data

  • 期刊论文

  • 发表日期：

    2026年6月22日

查看全文
  • Sahand Sobhi,
  • Rahim Ali Abbaspour,
  • Alireza Chehreghan

摘要

Numerous studies have evaluated the quality of Volunteered Geographic Information (VGI) using both extrinsic and intrinsic methods. This paper presents a new approach that combines both evaluations. The method introduces a feature matching technique that uses fuzzy normalized values of three geometric criteria: Hausdorff Distance, Direction Difference, and Area of Buffers’ Overlap. The study analyzes OpenStreetMap (OSM) history data for the study area and assesses the positional accuracy of the initial and most recent versions of OSM features by comparing them with an official reference dataset. As a primary objective, contribution volume, the number of individual contributors, and population density are examined as potential indicators of data quality. The analysis of contribution volume is carried out in four steps. First, comparing the initial and latest versions of features shows a significant improvement in positional accuracy, suggesting that repeated edits help refine data over time. Second, when features are grouped by usage type, accuracy improvements are observed, although the degree of change differs across categories. Third, analyzing features by their number of versions reveals a positive relationship between revision count and accuracy growth. Fourth, results from the latest dataset indicate a threshold effect, where additional revisions do not always lead to further improvement. In contrast, contributor count does not demonstrate a statistically significant association with positional accuracy, challenging the assumption that a higher number of contributors necessarily improves quality. Similarly, features in lower-density areas show slightly higher accuracy, suggesting that contextual factors may outweigh population concentration effects.