Recognizing Interest Groups in News Articles About Rulemaking

Zhou, Xiaoran

Permalink

https://hdl.handle.net/2142/128305 Copy

Description

Title

Recognizing Interest Groups in News Articles About Rulemaking

Author(s)

Zhou, Xiaoran

Contributor(s)

• Heng, Zheng
• Schneider, Jodi

Issue Date

2024-04-24

Keyword(s)

• Named Entity Recognition
• Text Mining
• News Analysis
• Rulemaking

Date of Ingest

2025-05-15T14:24:14-05:00

Abstract

58% of Americans oppose the U.S. Environmental Protection Agency’s new emission rules, and significantly more Republicans (83%) oppose these rules than Democrats (35%), according to the Pew Research Center. This project investigates whether news outlets often read by Republicans vs. Democrats mentioned different interest groups (such as the American Fuel & Petrochemical Manufacturers or the International Council on Clean Transportation) when reporting on emission rulemaking. We developed a Named Entity Recognition pipeline to identify the interest groups mentioned in the news. We collected 2,954 news articles that covered the emissions rule and so far manually annotated 45/2954 (1.52%) randomly selected news articles to evaluate our pipeline. The pipeline currently has an accuracy of 83.78% in recognizing interest groups in the news. In future work, we will annotate another 50 randomly selected news articles so that we can measure accuracy with a larger sample. We will apply the pipeline to recognize the interest groups in the 2,954 news articles. Results of this project will be used in the future to understand how interest groups and their opinions are presented differently by news outlets often read by Republicans vs. Democrats.

Type of Resource

Slides

Owning Collections