Analyzing positional accuracy of pois using spatial association rule learning

Friday 5:10pm, Main hall

Theme:
GIS, data use and data analysis

Audience:
(2a) Data users: Commercial

Alireza Kashian 30min

Center for Disaster Manangement @ University of Melbourne

We developed a new approach and associated methods to automatically validate and control for the positional accuracy of POIs in OSM. Thousands of POIs are added daily to OSM, while humand mistakes or vandalism activies can impact the quality of OSM data and hence lead to lower level of trust in data. Based on Tobler geography law, we assume nearby objects might have stronger relationships with each other, so a new analytic tool is developed to discover the association co-existence rules between each pair of objects such as (ATM,Building) or (Ferry Terminal, River). The tool is called OSM POI analyzer and accessible at www.openstreetmap.me. Using the tool, the user can select one target city and then add a sample POI (from the list) in any location he/she wishes. The background processor, will analyze all object pairs in 1000m radius and generates analytical reports which helps editos and OSM fans to understand the core layer of relationship between POIs and other sorrounding objects. The tool reveals how different urban designs will lead into different co-existence relations.