Wildlife trafficking is a growing illicit industry that threatens the survival of many species and is a major area of concern for law enforcement agencies and conservation groups globally. However, the lack of access to centralized aggregated databases on wildlife crime present challenges to experts seeking to understand and combat illegal trade. In this project, we developed a custom named entity recognition (NER) model to extract information about wildlife trafficking incidents from online English-language Indian news articles. By experimenting with different types of NER models, we trained a model that can identify trafficking-related entities with 94% F1-scoring. This is an improvement over a previous model developed for this task, which identified entities with 20% F1-scoring. We developed several data visualizations, which are available on our website for users to explore the extracted data.

• This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator

• Geary, Michael
• Andrade, Julie
• Huynh, Bao

Publisher

• Worcester Polytechnic Institute

Identifier

• E-project-032323-125455
• 94556

Palabra Clave

• Machine Learning
• Web Scraping
• Natural Language Processing
• Data Visualization
• Wildlife Crime

Advisor

• Konrad, Renata
• Lee, Kyumin
• Mangoubi, Oren Rami

Year

• 2023

Sponsor

• Dr. Patricia Raxter
• Dr. Meredith Gore

UN Sustainable Development Goals

• Not Specified

Date created

• 2023-03-23

Resource type

• Major Qualifying Project

Major

• Mathematical Sciences
• Computer Science

Source

• E-project-032323-125455

Rights statement

• In Copyright

Última modificación

• 2023-04-18