Uncovering Wildlife Intelligence with Taxonomy Management Tools in the Sensing Clues Foundation
Taxonomies and Ontologies
Sensing Clues is a non-profit organization dedicated to wildlife preservation. Since its conception in 2017, Sensing Clues has helped protect 1,202,540 hectares of nature across four continents in places like Brazil, Nepal, Burkina Faso, Northern Italy, Kenya, and the list goes on. The Semantic Web Company has been a very proud partner of Sensing Clues for three years now, where the PoolParty Semantic Suite has been a fundamental software to help realize Sensing Clues’ goals.
The Sensing Clues’ approach is built on a suite of tools that tracks movement and patterns in areas where illegal activity occurs, in order to hopefully curb the activity or prevent it based on the data that is collected. The suite of tools are deployed in the form of mobile and desktop applications which are put in the hands of field rangers responsible for safeguarding the wildlife surroundings.
In order to provide their services, Sensing Clues requires a strong taxonomy-based solution so that they can take advantage of synonyms, multilingual labels for their global clients, and contextual modelling in the taxonomy management tools.
“Five years ago I was in Laos…in a big forest reserve. In the middle of the night, I heard a gunshot”
And so began Sensing Clues, a non-profit foundation dedicated to the preservation of nature and wildlife through the use of deploying advanced technologies in the field. Founder and Director Jan Kees Schakel considers the gunshot he heard on vacation with his wife the impetus for starting the foundation because he realized that this “problem [illegal poaching] is actually something that requires a very professional approach.”
About the Sensing Clues tools & general workflow
Before we can understand how PoolParty Semantic Suite helps bring the Sensing Clues vision to life, it’s first important to understand how the foundation uses its applications to gather wildlife intelligence.
The Wildlife Tool Suite (WITS) comprises the following:
- Cluey: the data collector and tracking app used by the field rangers
- Focus Data Explorer: a data analysis tool that lets users access real-time observations made by the rangers, volunteers, camera-traps, etc. in order to get insights and statistics about the field activity
- Wildlife Conservation Analytical Toolbox (WildCAT): an advanced analysis and reporting tool that gives users occurrence likelihood and output maps based on heat maps and graphs
Photo courtesy of sensingclues.org/wits
Cluey serves as the foundational tool for Sensing Clues, where all data and insights are derived from the initial inputs keyed in by the field rangers. To give an idea of how Cluey and the remaining WITS tools are used, here is a simplified use case for a hypothetical field organization in Kenya:
- The Kenyan field organization has increasing problems with illegal poaching in the area, and also notes that habitats are being destroyed by human activity
- To combat these problems, the field organization partners with Sensing Clues and trains its field rangers how to use the technology
- Field rangers are supplied with handheld devices containing the Cluey app. On their daily patrols, the rangers record activity in the app which can range from animal sightings, to human activity (e.g. bushfires), to natural occurrences (e.g. area floods), to infrastructure changes (e.g. a broken fence), and the list goes on.
- The data entered by the rangers in Cluey serve as the basis for the statistics and patterns that can be explored in the Focus app through maps, timelines, and graphs of the raw data
- On top of the raw insights derived from Focus, the field organization can take it a step further with WildCAT. In this app, the field sightings are complemented with information derived from satellite images and other sources, as input for the calculation of predictive risk maps. These maps are used to anticipate illegal action and thus prevent it.
Where does the PoolParty Semantic Suite fit into this wildlife intelligence application? Through taxonomy management tools and ontologies.
The functionalities within the Cluey app are built primarily on PoolParty’s Thesaurus Server in order to take advantage of controlled vocabularies, synonyms, multilingual labels, contextual modelling, and more. From the backend, the Cluey developers add concept schemes and concepts to the taxonomy to model possible actions taken by field rangers, the observations they made, the typical animals within that field’s ecosystem, etc.
If we take the top level concept “Observation,” for example, the list of concepts can be narrowed to Human Activity, Animals > Fish, Mammals, Birds, and so on.
The concept Lion has a number of additional attributes such as synonyms, alternative labels, etc. Here, developers can add multilingual names for Lion under preferred labels, nicknames used in the field such as “King,” and also add linked open data for supplementary information pertaining to the lion’s biographical descriptions, its latin or scientific names, its photo, and more.
The key to this taxonomy is that it only models what is used in the field so that rangers using the Cluey app are familiar with the terminology. For this reason, PoolParty’s multilingual capabilities are especially important for rangers to be able to use their native languages, and the makeup of a semantic concept is necessary to provide additional descriptive and contextual information about that concept so that rangers are not limited to one attribute. In other words, if Cluey was only built off of simple terms, Lion would just be Lion; as a concept, Lion is Lion, Leão, Lioness, King, Panthera Leo.
From the frontend, Cluey mimics precisely this.
As a ranger clicks within Cluey, the available options to key in for reporting are narrowed based on the taxonomy’s hierarchy in the backend. Additionally, the PoolParty ontology helps express relations between the concept schemes of the taxonomy to model the actions that the ranger takes. So, a ranger who is on foot patrol (A) spots a lion (B) and scares away (C) the lion from antelope in the area that come near a school.
Photo courtesy of sensingclues.org/cluey
The ranger can track their movements and observations in the app in order to feed the Focus and WildCAT applications for analysis and predictive modelling. In fact, the user input fields within an observation report largely resembles that of PoolParty’s taxonomy management tools, where by toggling different concepts, this data combined with the help of analytic software can produce an observation risk map as pictured on the right.
Photo courtesy of sensingclues.org/cluey
The taxonomy and ontology capabilities in the PoolParty Thesaurus Server are the backbone of the user-friendly applications that Sensing Clues delivers to organizations in the field.
The defining characteristics of PoolParty’s taxonomy management software.
When Founder Schakel set out to find software partners, he had a few important things in mind that he knew needed to be addressed. According to Schakel, the software needed to have sophisticated taxonomy management tools to meet the demands of the work, but easy to use in order to train the foundation’s volunteers who range from data scientists to software engineers to biologists.
Considering the complexity of semantic technologies, “easy-to-use” is not always synonymous with the field. However, PoolParty Semantic Suite delivers advanced technology in a user-friendly interface that easily integrates with existing databases or systems.
As Schakel put it, “I wanted a tool that would help me model the field of biodiversity and law enforcement within the biodiversity domain…[and with PoolParty] I could use it from the very start, almost without any help.”
This way, Schakel could start the foundational modelling for the Sensing Clues taxonomies, and volunteers from his team could easily contribute their own expertise to it without much training hassle.
PoolParty Semantic Suite makes modelling the domains quite easy and helps cement the capabilities within the Sensing Clues Wildlife Tool Suite much quicker. In order to serve their field organizations with easy-to-use, understandable solutions, the software must use a “pragmatic” thesaurus that is “recognizable” and “translatable,” which PoolParty’s straightforward user interface helps facilitate.
The Sensing Clues – PoolParty partnership, going forward.
As Sensing Clues continues to deliver quality services to their field organizations in need, they also expect to expand WITS’ offerings. They want to build on top of the data visualization tools they already have by making use of the Sensing Clues knowledge graphs in PoolParty.
Schakel and his team are working on ways to show the field rangers how they can interpret relations between concepts of the thesaurus to spot patterns in the data. In following Sensing Clues’ key principle, these graph visualizations should be intuitive for their rangers to follow, so they are testing and designing different models of the knowledge graphs to ensure this is the result.
In the meantime, Sensing Clues remains fully dedicated to their mission and field organizations who are benefitting from such innovative solutions, which is proudly supported by PoolParty Semantic Suite’s Thesaurus Server.
The Story at a Glance
Want to learn more about Sensing Clues? Head over to their website for more information.
The PoolParty Semantic Suite specializes in taxonomy and ontology management solutions. Watch our Taxonomy Management 101 webinar with taxonomy expert, Heather Hedden, to hear the benefits and see demos.