This post was co-authored with Kevin Lynch.
In October, BioBlitz 2011 took place in Tucson’s Saguaro National Park East and West. Thousands of volunteers worked together to discover the biodiversity of this marvelous place I call home. This blog entry outlines the work we’ve done the last few months, the reasons why BioBlitz matters (they might surprise you), and makes a call to photographers to help us test our crowdsourced image classification process.
The Team – National Geographic, National Park Service, Encyclopedia of Life, National Park Foundation
People from around the country worked hard to make BioBlitz successful. There were – and still are – a lot of moving parts. The Park is over 100 square miles, 70% of which is officially “wilderness,” which means, among other things, no wheels allowed! The National Park Service coordinated all efforts on the ground. National Geographic co-sponsored, and provided FieldScope, a geographical information system customized specifically for BioBlitz. The National Park Foundation had an electronic field trip on the first day of BioBlitz, complete with a companion website, lesson plans, and games. The Encyclopedia of Life provided their species taxonomy, a BioBlitz-focused community, and has 8 featured collections for BioBlitz, including a list of all the species we expected to find in the Park that do not yet have images.
What we’ve done
We at TriviumRLG began talking about the possibilities for using semantic technology to organize and classify images for BioBlitz earlier this year at South by Southwest Interactive (here) and the Semantic Technology Conference (here). The idea is really simple: crowdsource the image classification process, using free software (Microsoft’s Pivot, which we’ve written about). Do it quickly at low cost. If someone takes a photo and tags it with the word “butterfly,” we’ll attach the kingdom, phylum, genus, and species for “butterfly” from the Encyclopedia of Life species taxonomy. Or, if all someone knows is that they have a picture of an insect, we’ll get the taxonomy information as far as we can. Sure, people will get it wrong sometimes. But with a simple combination of volunteers, scientists, and a comprehensive taxonomy the scientists understand, image classification errors can be caught and fixed very quickly. One correct tag on a photo allows us to add MANY more facets of information. For example: the original tag on this photo was:
From that we were able to lookup and add:
This just-in-time lookup provided several benefits for the project team, the volunteers on the ground, the scientists who will use this data for research, and the conservationists who will identify and prioritize activities in Saguaro National Park.
The suite of tools provided by the team enabled faster time to classification of information, up-to-date visualizations of activity on the ground, clearer understanding of gaps in the information and how to route the incomplete objects to the best expert to finalize and approve them for use in several known and many yet-to-be-known applications.
Why it matters
We can’t protect something if we don’t know it’s there. We have incredible biodiversity in our back yards, and we proved it at BioBlitz. From birds to bacteria, people took pictures, and we helped classify them, quickly and cheaply. From a semantic technology standpoint, we did something very simple, but the fact we did it over 100 square miles in a single 24-hour period with volunteers and zero investment and had instant, meaningful results makes this important. We engaged in citizen science and created a model other BioBlitzes can use. Professionally, BioBlitz has been an incredible opportunity to contribute to our field. Personally, we feel that we have made a difference, for the world, and for our children.
- Smart TVs and the Semantic Web
- A Tale of Agile Development... of a Standard
- Tagging the Visual Web: Visual Media Doesn't Have To Be Dumb Anymore
- Introduction to: Open World Assumption vs Closed World Assumption