An AI-Based Approach To Identify Spatio-Temporal Contribution Patterns And Their Impact On OpenStreetMap Data Quality

This work enters the field of Geospatial Information Technology, targeting the lack of spatial data prevalent in developing regions.Voluntary Geographic Information (VGI) has become one of the major research topics in GIScience, more than fifteen years after the term was coined by Goodchild (2007).In many developing countries, where spatial data deficits are common at both large and small scales, Silva and Camboim (2020) suggest that VGI can mitigate the consequences of this scenario.Utilizing VGI via OpenStreetMap (OSM), the study aims to enhance urban planning efforts in Brazil.OpenStreetMap, with over 7.8 million registered users and more than 16.7 billion edited features, is recognized as one of the world's largest and most popular platforms for collaborative data collection Teimoory, Abbaspour, and Chehreghan (2021).OSM has a significant advantage over commercial geodata providers because it adheres to the Open Data Commons Open Database License, which grants unrestricted data availability for a wide range of applications, promoting greater usability and flexibility.Researchers such as Brovelli and Zamboni (2018), Nasiri et al. (2018), and Haklay ( 2010) have focused their efforts on understanding the characteristics of OSM, particularly in terms of data quality and feasibility in integration processes.Understanding this is crucial for effectively using collaborative datasets like OSM.ISO 19157 (ISO, 2013) is the prevailing global standard for assessing spatial data quality, including completeness, thematic accuracy, logical consistency, temporal quality, positional accuracy, and usability elements.In 2015, the Geographic Service Directorate in Brazil created the Technical Specification for Geospatial Data Quality Control (ET-CQDG).ET-CQDG lays out criteria based on the ISO 19157 standard for evaluating geospatial data quality.Furthermore, it presents various sophisticated sampling techniques to support the essential analyses.To complement these guidelines, in 2019, the Brazilian Institute of Geography and Statistics published the Technical Manual for Geospatial Data Quality Assessment, which has provided valuable standards and methods for evaluating the quality of this data.While official agency maps exhibit uniformity due to professional expertise, VGI platforms do not guarantee this consistency.Consequently, global efforts focus on researching and analyzing VGI data quality parameters to assess their integration into official reference bases.Traditional measures often rely on comparisons with official data.However, Brovelli and Zamboni (2018) suggest using intrinsic parameters as an alternative approach to assess VGI quality.Mooney, Corcoran, and Winstanley (2010) support this, indicating intrinsic parameters' utility without comparative methods.Barron, Neis, and Zipf (2014) identified intrinsic indicators as feasible when quality is not strictly tied to official evaluation parameters.Grinberger et al. ( 2021) and Brckner et al. ( 2021) conducted mathematical modeling studies in OSM, analyzing contributions over time to detect global events and estimating the completeness of OSM retail stores in Germany, respectively.Building on VGI, this research employs logistic regression models to analyze the completeness and growth dynamics of urban mapping within OSM.By collecting historical OSM data through the OSHOME API, the study applies logistic regression to model mapping activities' temporal evolution.This model identifies different mapping stages-initial growth, expansion, and saturation-providing a detailed urban mapping assessment.The core objective is to use AI to systematically investigate spatio-temporal contribution patterns and their impact on OSM's geospatial information quality.This involves: (1) collecting OSM data for Brazilian cities from 2008 to 2023, (2) preprocessing the data using Python libraries like Pandas and Geopandas, (3) applying a four-parameter logistic regression model to the cumulative contribution history, and (4) extracting and analyzing logistic parameters to categorize cities based on mapping growth dynamics.