:: Volume 6, Issue 1 (10-2016) ::
JGST 2016, 6(1): 19-32 Back to browse issues page
Introducing a Conceptual Model to Improve the Quality of Storage of Volunteered Geographic Information: In the Field of "Fitness-for-Use" Indicator
M. Eshghi *, A. A. Alesheikh
Abstract:   (3432 Views)

Quality assessment is one of the most important challenges on the way of Volunteered Geographic Information – VGI. Nowadays, assessing VGI data quality mainly includes documenting and measuring errors with no concern about satisfying various needs of users. Despite the widespread acceptance of “Fitness-for-Use” as one of the key quality elements of VGI data, it is not considered practically in producing as well as sharing of VGI data. Therefore, the aim of this paper is to present a conceptual model so as to consider this indicator as a key to better storage of data. Geographical features and the acquired data from them have different applications; for instance, a road could have roles in both Navigation and Emergency Management applications. With this in mind, various kind of data applications would be considered by different users at the time of using them according to their needs. What is more, it is possible that different applications of data at the time of producing as well as storing data have been considered by different users. Consequently, an acquired data about an object may be stored more than once with different attributes (labels related to applications), or no consideration of other possible applications of data. This will result to redundancy and incomplete attribute of those data, respectively. Therefore, in this paper by designing a conceptual model – which is inspired by CityGML concepts – it is tried to apply some preparation in the process of storing this type of data as comprehensive as possible, of course without inconsistency, so as to make an opportunity for better as well as optimal use of them. This is done by considering a set of labels which represent roles of a specific data in different applications. By doing so, if a user just considers a single application of a data at the time of storing, other application will automatically detect and attached to the data. For example, a single linear object could be considered as Primary road, Arterial road as well as Highway. As a result, data storage will be improved, which can be led to better processing and better presenting data as a response to the  users’ queries – needs and purposes. To implement the proposed conceptual model, the Zhao’s semantic similarity measurement algorithm is used. OpenStreetMap is used as the source of VGI data: linear data including Streets and Roads of districts 2, 5 and 9 of Tehran municipality are selected for this paper. ArcObject programming is used in C#. SQLSERVER is used as the DBMS, and ArcMap is used for presenting the results. According to the proposed conceptual model existing data as well as new imported data are processed and stored. Therefore, many application could be considered for a single data which will result a more complete data. In the following, in order to evaluate the performance of the presented model, database response to the queries of the users is analyzed and compared in two different databases: one with ordinary available storage type, and the other one with new storage type presented in this paper. The results show that compared to the ordinary database, responses by the new database to a group of users are: 1) more available in different queries, and 2) more accurate toward the user’s needs and purposes. The improvement of the performance of the new database is also presented based on the number of correct answers to 15 queries by 10 different users, which is more than 50 percent. As a recommendation for future works, the proposed model can be used for creating an automatic filter at the time of data entry.

Keywords: Volunteered Geographic Information (VGI), Data Quality, Fitness for Use
Full-Text [PDF 1356 kb]   (1594 Downloads)    
Type of Study: Research | Subject: GIS


XML   Persian Abstract   Print



Volume 6, Issue 1 (10-2016) Back to browse issues page