December 18, 2018 - Dale Kim | Big Data Ecosystem

Geospatial Analytics at Big Data Scale and Speed

You look at maps all the time these days, especially as part of your Internet searches. Whether you seek directions to a new restaurant, current traffic to the airport, or home prices in your area, you get better context and much more complete answers to questions when maps are involved.

If you don’t currently use maps as part of your daily analytics work at the office, you probably will soon. After all, more business decisions today across all industries are leveraging geographical information, and mapping technologies integrated into your business intelligence (BI) platform help tell the whole story. In many situations, standard BI charts won’t give you enough context. For example, financial services firms might use geospatial data in conjunction with other forms of alternative data to track down financial criminals as part of their regulatory compliance requirements. Or, as a retailer, you might want to assess new store locations based on average residential income in the surrounding area. Traditional visuals can show you how nearby regions compare to one another, but that same data is so much more powerful when displayed as a choropleth map. Such a visual would make it easy to capture more information relevant to your decision, including what other stores are in the desired areas and what infrastructure would enable high traffic to your store. Accompanying that choropleth map with other BI charts that show, for example, the relationship between neighboring residential income and store revenues in your other existing locations will give you a better sense of what locations might be ideal.

Mapping Technologies Expect Big Data
The theory behind the scenario described above is very straightforward, but there are challenges to the actual practice. For one, it is extremely important to look at your data at all levels of granularity so you’re not missing out on critical details. This means your system must be able to handle a lot of data. Using less data means less detail, so instead of viewing individual house prices or average house prices at a block level, you end up viewing average house prices at a higher level of detail like zip code or county. Second, not all mapping technologies were designed for dynamic updates as the underlying data gets updated. We typically think of maps as a static visual, but data continues to change, and you want to make sure your maps reflect the latest data, down to the very millisecond. Related to this is how real-time analytics will fit in with your mapping visualizations. With the growing interest in real-time analytics, especially on streaming data, you must ensure that your geospatial analytics platform is ready for fast analysis on constantly changing data. And finally, you want the flexibility to define custom geospatial shapes that allow you to specify your own boundaries and not be limited to a predefined set of boundaries that might not apply to your business.

In today’s mapping systems, the challenge doesn’t lie with the mapping technologies themselves, but rather, in the analytics platform that serves up the data. This is where BI platforms native to big data like Arcadia Data can bring out the most in mapping technologies. Our integration with Mapbox GL lets you take advantage of big data stores and present geospatial visualizations without the limitations found in traditional BI tools that incorporate mapping. Arcadia Data can easily handle billions of records stored in your data lake that represent millions of shapes so you can view an extreme level of granularity. The maps calculations are done on the server side, which enable fast re-rendering of maps when the underlying data changes. This capability also lets you change metrics on live data and get an immediate update without the long delays typical of other systems. And you can use Mapbox Studio to upload custom shape files for business-specific geospatial boundaries you can analyze.

See for Yourself
If you are interested in trying this out yourself, we will soon have this capability in our freely downloadable Arcadia Instant product, a desktop BI tool that lets you explore our visualization capabilities. Or you can contact us now, and we will walk you through a demo or work with you on setting up a trial version of our Arcadia Enterprise product.


Related Posts