A Geospatial Intelligence System for Land Price Modelling and Property Management: An Offline/Online Android Application for Nyamata Sector, Rwanda
East African Journal of Information Technology · 20.11.2026 · 9 (3) · 252-269
DOI: 10.37284/eajit.9.3.6343 ↗Kurz gefasst
Fachartikel in der Zeitschrift East African Journal of Information Technology, online erschienen am 20.11.2026. Verfasst von Jean Claude Nshimiyimana und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Property valuation and management in Rwanda face challenges such as fragmented data, manual valuation methods, inconsistent pricing, and limited use of geospatial technologies, leading to inaccurate valuations and inefficient land administration. To address these challenges, this study developed a Geospatial Intelligence System integrating GIS, Remote Sensing, Artificial Intelligence (AI), and an Android offline/online application for property management and land price modelling in Nyamata Sector, Bugesera District. Using a mixed-methods approach, the study integrated spatial and non-spatial data from cadastral records, satellite imagery, valuation databases, socio-economic datasets, and GPS/GNSS field surveys. Data were collected through satellite imagery, field observations, questionnaires, interviews, and institutional records. A purposive sample of 105 stakeholders, including land officers, valuers, GIS specialists, planners, regulators, telecom professionals, and property owners, participated in the study. Advanced analytical tools such as ArcGIS, QGIS, Python, Multiple Linear Regression, Geographically Weighted Regression, Random Forest, Neural Networks, and XGBoost were used for analysis. The results showed successful integration of over 85% of datasets, with more than 90% of land parcels linked to thematic information. The AI-based valuation model demonstrated excellent predictive performance (R²=0.985; RMSE=222,921 RWF) and achieved over 90% agreement between predicted and actual market values, with an overall spatial accuracy of 93.8%. Accessibility, digital infrastructure, environmental quality, and neighbourhood conditions were identified as the main factors influencing land values. The Android application effectively supported automated valuation, parcel identification, GIS mapping, and decision-making. The study concludes that geospatial intelligence significantly improves valuation accuracy, transparency, and efficiency in property management. It recommends nationwide adoption of GIS- and AI-based valuation systems, enhanced geospatial data sharing, and integration of real-time market data, drone imagery, and advanced AI models to strengthen sustainable land governance in Rwanda.
Zitieren
APA
Nshimiyimana, J. C., Nsanziyera, A. F., Idukunda, C., & Hagenimana, E. (2026). A Geospatial Intelligence System for Land Price Modelling and Property Management: An Offline/Online Android Application for Nyamata Sector, Rwanda. East African Journal of Information Technology, 9(3), 252-269. https://doi.org/10.37284/eajit.9.3.6343
Vancouver
Nshimiyimana JC, Nsanziyera AF, Idukunda C, Hagenimana E. A Geospatial Intelligence System for Land Price Modelling and Property Management: An Offline/Online Android Application for Nyamata Sector, Rwanda. East African Journal of Information Technology. 2026;9(3):252-269. doi:10.37284/eajit.9.3.6343
BibTeX
@article{Nshimiyimana2026,
author = {Nshimiyimana, Jean Claude and Nsanziyera, Ange Felix and Idukunda, Clémence and Hagenimana, Emmanuel},
title = {A Geospatial Intelligence System for Land Price Modelling and Property Management: An Offline/Online Android Application for Nyamata Sector, Rwanda},
journal = {East African Journal of Information Technology},
year = {2026},
volume = {9},
number = {3},
pages = {252--269},
doi = {10.37284/eajit.9.3.6343}
}
EUVIMED