2023 Undergraduate Research Showcase


Human-Wildlife Conflict

Document Type

Student Presentation

Presentation Date


Faculty Sponsor

Dr. Matt Williamson


Conflicts between people and wildlife can result in monetary loss, agricultural damage or even injury and mortality of a person or the wildlife. Reducing these conflicts is difficult due to differences of opinion regarding the importance of wildlife and the appropriateness of the strategies we use to manage them. Understanding common themes that underlie conflict could lead to better management practices for agencies and civilians alike and highlight shared perceptions of the benefits that wildlife can bring. We are hoping to identify the range of conflicts and better understand public opinions by combining machine learning and article coding to look at thousands of articles. Using the NewsBank database, we found over 150 different articles from around the country that describe interactions between humans and wildlife and performed content analysis in order to categorize the themes within the articles. We focused on a select group of species, consisting of grizzly bears (Ursus arctos), wild boars (Sus scrofa), and beavers (Genus castor). Through the reading and content analysis we were able to build a codebook consisting of 20 different coding themes. Preliminary results show a clear indication that the discourse around human-wildlife interactions tends to be negative, regardless of species type. Although this is especially true within the species of Boars and Grizzly Bears, we also found that Beavers have a little more of a positive sentiment analysis. The type of negative outcome can vary drastically, depending on species and interaction type, but the fact remains that in general, humans and wildlife struggle to cohabitate peacefully. Further research will include; 20 continual coding of articles, possible expansion of the codebook, the inclusion of more species residing in North America in order to determine different conflicts and outcomes, and using the coded article corpus to build and train a supervised classification algorithm. The benefits of our approach can create safer environments for wildlife and humans. We anticipate that these findings and analyses can help more effectively manage conflicts and compromise with opposing parties.

This document is currently not available here.