Data Quality in Software as a Service Implementation of Public Geographic Information System

Public Geographic Information system (GIS) is a system that automates the process of collecting community contributions of spatial data cross referenced to base maps. It provides the capabilities of analyzing such spatial data giving effective decision support information such as decisions for appropriate land use planning, planning pedestrian areas in crowded streets. On the other hand, the easiest and lowest cost of ownership way to let people to consume public GIS services is to provide those services over cloud using software as a service (SaaS) model. The probability of introducing low quality data over base map through community contributions represents a major drawback in SaaS implementation model for public GIS. This paperʹs objective is to overcome such drawback, by introducing new method for map layerʹs content verification. Which depends on five factors, voting on map-layers; user complains as per user type individual, group user /corporate; existence of agreed Service Level Agreement (SLA), and reputation of spatial dataʹs author. Based on a combination of these factors data quality was measured so that decisions for adding, merging new or cross referencing map-layers can be taken based on predefined percentage of measured-data-quality; hence data quality of new map data presented to users could be enhanced. A reference data quality percentage of 0.75 was proposed to help taking decision confidently for adding, merging or cross referencing new map layers with base map; The quality scoring for sample map layers was extracted from running public GIS in private cloud using the proposed factor over four contributed map layers, which are accessed by 77 users, and the result was two layers passed with data quality percentage of 77.5%, and 75.9% respectively. A reliable method was introduced in SaaS implemented public GIS to approve new map layers with reasonable quality contributed by community.