GeoDescriber: Expanding the Reach of Geographic Information

Landscape description of Whanganui National Park, New Zealand.
Photo credit: James Shook. Text credit: GeoDescriber

Can your GIS talk?

Does it say the right things?

In the era of too much information, demand for a smart solution to automatically interpret data for a specific audience is rising. GIS is no exception. Maps help us understand the geography, but more is possible. That’s why I’m excited about something new we’ve been working on. We named it GeoDescriber.

At its simplest, GeoDescriber is GIS tool that writes prose to describe places. Behind the scenes, it analyzes the wealth of information in the Living Atlas of the World and describes places starting with what’s most important to know.

GeoDescriber was written to describe landscapes in a useful compelling way, like a regional geographer might. You specify places (in a feature class), and GeoDescriber will characterize the landscape within those polygons. To see examples, see our story map called Introducing GeoDescriber.

GeoDescriber is free, and available on GitHub. By default, GeoDescriber characterizes landscapes. However, GeoDescriber can be modified to describe other phenomena, such as prevailing political attitudes or customer profiles. GeoDescriber only needs to be pointed towards the right data.

Westgard Pass,
Photo Credit: Marshal Hedin – Westgard Pass Rd, California. Text credit: GeoDescriber

Among many possible applications for this technology, I think these are particularly interesting:

Drop GeoDescriber’s text into a text-to-speech synthesizer. Then, create a map linking places with the sound files of the descriptions. There is potential here for a new kind of map for the visually impaired, and for applications which address section 508 accessibility requirements.

GPS applications tell you when to exit the highway. But perhaps you could make a GPS application that tells you about the landscape you see along the way? To do this, run GeoDescriber on viewsheds generated along a road.

About the author

Michael Dangermond has over 25 years experience in GIS in such diverse fields as cartography, agriculture, international boundary delineation, environmental protection, regional planning, park planning, land and wildlife conservation, and forestry. He has been working for ESRI since 2010.

Notify of
Inline Feedbacks
View all comments

Next Article

Virtualizing 3D training models with NVIDIA AI Enterprise

Read this article