Assessment of VGI Data Quality from Fitness for Use Perspective in Accident Management, Case Study: Routing and Navigation

Suitable spatial data is one of the most important needs of managers and planners in the field of accident management. An example of such accident was the fire and collapse of PLASKO building, one of the commercial and shopping center in Tehran. Over the past few years, regard to development of technology such as social networks, cloud computing, the participation of non-professional public in collecting and sharing geographic information has been increased. These type of information calls volunteered geographic information (VGI) and could be a precious alternative for authoritative counterpart due to offering substantial characteristic like free use of data, local knowledge and up-to-date creation, although VGI suffers from a lack of quality awareness which has led to hesitate about usage of this type of data in this particular subject. Therefore, assessment of VGI quality is necessary in terms of fitness for use in accident management. In this article, we have chosen routing of relief group as one of the most important needs in the PLASKO case. Moreover we have selected fitness for use as the quality indicator not only it is dominant aspect of quality which depends on user's need and intended purpose but also, this quality indicator was discussed generally on the other scholars work. Therefore, we aim to propose a workflow which could support the decision whether VGI dataset could be adequate for accident management or they could meet the active group's expectations or not. In order to perform, geometric and descriptive information was identified at first. For example Edges and Nodes of road network are the required geometric information. Name, speed, type, one way-two way and so on are descriptive information of road network. From the descriptive information, we have assessed name of road network. Then quality indicators accordance to mentioned information should be determined. Completeness, positional accuracy, logical consistency applied as quality indicators for describing edges and nodes, attribute accuracy and attribute completeness used as quality indicators of roads' name. Next, we have estimated the indicators by using intrinsic or extrinsic ways. Besides, we have applied a grid-base approach to divide the region and find the tiles with high heterogeneous or worse quality. After all, In order to assess fitness for use, which is described by other indicators, a fuzzy expert system was used to aggregate entire indicators. In the implementation process, district 11 of Tehran was selected as the study region. OSM site was used as VGI data set. Also, the data generated by the municipality was used as the standard data set. Meanwhile, in this article, a grass-based module was adopted to match two datasets as a prerequisite step in extrinsic ways. The results of this paper indicates that 74.4% of the total area has a fitness for use (0.93- 0.9) and the rest of all are in the range of (0.5-0.91). On average, the whole region has 89% fitness for use. In the following, analysis of the shortest path as a criteria for evaluation was applied in two datasets. The result was shown 95% similarity of OSM data compared with standard data.