GIS and NoSQL Databases

| |


Traditionally we see GIS as a form of database that encapsulates spatial information in a relational structure. Increasingly, however, researchers and industry are realizing that relational databases present some limitations to scaling, preventing or limiting the applicability of big data and real-time data problems utilized within GIS. Large volumes of data accessed quickly require more efficient methods for structuring and presenting data to a spatial platforms in real-time, requiring alternative and easier methods for database integration.[1] One reason why NoSQL data are now becoming popular are the vast quantities of unstructured or semi-structured data generated by popular social media platforms and websites. This has led to greater needs for new data structure formats for databases, including those that provide alternative paradigms such as object-oriented models or, as the name suggests, do not require the use of SQL queries that can be complex. As an example, retrieving data within objects embedded with XML or markup languages allows one to use object inheritance or interfaces that give far greater flexibility for data formats.

Most proprietary GIS tools are still utilizing traditional relational formats; however, as GIS begins to adopt NoSQL databases, many still retain SQL-like queries. One popular tool includes MongoDB, which is a NoSQL database that is open source and can be integrated with many open source tools such as QGIS. Data retrieval is also generally faster with NoSQL formats and is often easier to structure within existing code for programs, not requiring complex SQL queries that could be less efficient for given data problems. Although there is still hesitancy among leading industry companies to fully adopt NoSQL structures, among new and emerging companies this has become more of the norm given increased demand to deliver fast output for queries. NoSQL also facilitates analysis and integration within a variety of tools, which is why open source GIS has proven to be the most useful arena for NoSQL databases. Not surprisingly, web-based GIS is probably the area that is currently leading in the use of NoSQL databases within GIS, as types of real-time data are more typically found in these platforms.[2]

Query result of GTS spatial data in mongoDB. From: Zhang, Song, & Liu, 2014.
Query result of GTS spatial data in mongoDB. From: Zhang, Song, & Liu, 2014.


[1] For information and examples on the advantages of a NoSQL structure for GIS applications, see:  Zhang, X., Song, W., & Liu, L. (2014). An implementation approach to store GIS spatial data on NoSQL database (pp. 1–5). IEEE.

[2] For an example of a NoSQL GIS tool and project, see:  Liu, Jin; Wu, Yangping; Zhuo, Dan; Hu, Qiping; Wang, Yan; et al. International Journal of Digital Content Technology and its Applications 6.20 (Nov 2012): 268-276.

See Also


GIS and Natural Language Processing

Using an Interactive Web and Desktop Based Spatial Technology to Monitor Performance of Public Health Outcome in Developing Countries