Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Bike Frames: Understanding the Implicit Portrayal of Cyclists in the News

Xingmeng Zhao, Dan Schumacher, Sashank Nalluri, Xavier Walton, Suhana Shrestha, Anthony Rios

TL;DR

This paper tackles how news framing shapes public perception of cyclists by introducing the Bike Frames dataset and the BikeFrame Chain-of-Code prompting framework. The approach uses pseudocode-driven reasoning to jointly predict accident mention, fault attribution, and perception while incorporating news bias analysis to improve predictions. Empirical results show substantial improvements in perception and fault classification when bias information and self-consistency are included, with an average F1 gain from .739 to .815. A US-focused case study uncovers differences in reporting across outlet types and gender-related framing, highlighting the influence of media on cycling safety perception and infrastructure support. The work offers a practical framework for bias-aware news analysis and informs design of tools to promote balanced public discourse around cycling and road safety.

Abstract

Increasing cycling for transportation or recreation can boost health and reduce the environmental impacts of vehicles. However, news agencies' ideologies and reporting styles often influence public perception of cycling. For example, if news agencies overly report cycling accidents, it may make people perceive cyclists as "dangerous," reducing the number of cyclists who opt to cycle. Additionally, a decline in cycling can result in less government funding for safe infrastructure. In this paper, we develop a method for detecting the perceived perception of cyclists within news headlines. We introduce a new dataset called ``Bike Frames'' to accomplish this. The dataset consists of 31,480 news headlines and 1,500 annotations. Our focus is on analyzing 11,385 headlines from the United States. We also introduce the BikeFrame Chain-of-Code framework to predict cyclist perception, identify accident-related headlines, and determine fault. This framework uses pseudocode for precise logic and integrates news agency bias analysis for improved predictions over traditional chain-of-thought reasoning in large language models. Our method substantially outperforms other methods, and most importantly, we find that incorporating news bias information substantially impacts performance, improving the average F1 from .739 to .815. Finally, we perform a comprehensive case study on US-based news headlines, finding reporting differences between news agencies and cycling-specific websites as well as differences in reporting depending on the gender of cyclists. WARNING: This paper contains descriptions of accidents and death.

Bike Frames: Understanding the Implicit Portrayal of Cyclists in the News

TL;DR

This paper tackles how news framing shapes public perception of cyclists by introducing the Bike Frames dataset and the BikeFrame Chain-of-Code prompting framework. The approach uses pseudocode-driven reasoning to jointly predict accident mention, fault attribution, and perception while incorporating news bias analysis to improve predictions. Empirical results show substantial improvements in perception and fault classification when bias information and self-consistency are included, with an average F1 gain from .739 to .815. A US-focused case study uncovers differences in reporting across outlet types and gender-related framing, highlighting the influence of media on cycling safety perception and infrastructure support. The work offers a practical framework for bias-aware news analysis and informs design of tools to promote balanced public discourse around cycling and road safety.

Abstract

Increasing cycling for transportation or recreation can boost health and reduce the environmental impacts of vehicles. However, news agencies' ideologies and reporting styles often influence public perception of cycling. For example, if news agencies overly report cycling accidents, it may make people perceive cyclists as "dangerous," reducing the number of cyclists who opt to cycle. Additionally, a decline in cycling can result in less government funding for safe infrastructure. In this paper, we develop a method for detecting the perceived perception of cyclists within news headlines. We introduce a new dataset called ``Bike Frames'' to accomplish this. The dataset consists of 31,480 news headlines and 1,500 annotations. Our focus is on analyzing 11,385 headlines from the United States. We also introduce the BikeFrame Chain-of-Code framework to predict cyclist perception, identify accident-related headlines, and determine fault. This framework uses pseudocode for precise logic and integrates news agency bias analysis for improved predictions over traditional chain-of-thought reasoning in large language models. Our method substantially outperforms other methods, and most importantly, we find that incorporating news bias information substantially impacts performance, improving the average F1 from .739 to .815. Finally, we perform a comprehensive case study on US-based news headlines, finding reporting differences between news agencies and cycling-specific websites as well as differences in reporting depending on the gender of cyclists. WARNING: This paper contains descriptions of accidents and death.
Paper Structure (14 sections, 3 equations, 8 figures, 5 tables)

This paper contains 14 sections, 3 equations, 8 figures, 5 tables.

Table of Contents

  1. INTRODUCTION
  2. RELATED WORK
  3. DATA COLLECTION AND ANNOTATION
  4. METHODOLOGY
  5. DESIGN: Pseudocode
  6. EXECUTE: Simulate Pseudocode Execution
  7. RESULTS AND DISCUSSION
  8. LIMITATIONS AND FUTURE WORK
  9. CONCLUSION
  10. Paper Checklist to be included in your paper
  11. Appendix
  12. A.1. Baseline Model Descriptions
  13. A.2. Summary Statistics for Male vs Female Articles for Unlabeled Corpus
  14. A.3. Prompts Used in Our Experiments

Figures (8)

  • Figure 1: Understanding and explaining implicit views of cyclists requires reasoning. Hence, our "Bike Frames" aim to (1.) detect accident-related headlines, (2.) identify implicit suggestions of "who was at fault", and (3.) measure perception towards cyclists.
  • Figure 2: BikeFrame Chain-of-Code prompt. The news information component is highlighted in yellow.
  • Figure 3: The proportion of accident-related headlines in the domain-specific (e.g., bicycling.com) and general news websites (e.g., nypost.com and chicagotribune.com).
  • Figure 4: The proportion of headlines in domain-specific websites vs. general news-based websites on who is perceived to be at-fault for the accident, the cyclists, or other entities in the headline (Other).
  • Figure 5: The proportion of male/female headlines unrelated to an accident for cyclists.
  • ...and 3 more figures