Analyzing Public Perception of Autonomous Vehicles Through Social Media Data: A YouTube Comment Study on Tesla Autopilot Accidents

Wang, Chenyu; He, Jiayi · 2026 · Crossref

DOI: 10.54254/2753-7048/2026.hz32369

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates public perception of autonomous vehicles by analyzing emotional and cognitive responses to Tesla Autopilot accidents on YouTube. Motivated by the limitations of existing research—which often relies on surveys, lab tests, or broad social media categories lacking specificity—the authors aim to capture the organic intensity of public discourse through a detailed case study. The research addresses three questions: identifying central thematic frames in YouTube comment sections, profiling the emotions associated with these frames, and determining how media reporting shapes prevailing discourses. The methodology involves extracting 437 English-language comments from three highly viewed YouTube videos covering Tesla Autopilot incidents between 2016 and 2025. The videos were selected for high view counts, variety in accident types (fatal and nonfatal), and diverse sources (mainstream and independent channels). The authors employed Latent Dirichlet Allocation (LDA) for topic modeling, identifying four distinct themes: Technological Skepticism, Satire and Comparison, Safety Concerns, and Attribution of Responsibility. Sentiment analysis was conducted using SENA software based on the NRC Emotion Lexicon to measure eight primary emotions. Additionally, Quadratic Assignment Procedure (QAP) analysis was used to compare emotional compositions across the three videos. The findings reveal four polarized themes with distinct emotional profiles. "Technological Skepticism" and "Safety Concerns" are dominated by negative emotions, particularly fear and sadness, reflecting anxiety over system reliability and life-threatening risks. "Satire and Comparison" features a complex mix of anger, disgust, anticipation, and surprise, indicating humorous or sarcastic engagement rather than genuine safety concern. "Attribution of Responsibility" stands out with higher levels of trust and rationality, as commenters argue for shared liability between drivers and the system. QAP analysis showed that emotional responses to two of the videos were statistically similar, focusing on safety fears, while the third video elicited significantly different emotions, shifting toward responsibility attribution and ridicule. The study concludes that public acceptance of autonomous driving is not static but dynamically mediated by emotional responses to specific incidents and media framing. The authors propose an "Incident-Emotional Response" paradigm, suggesting that trust decay and maintenance depend on transparent communication and addressing emotional concerns. For businesses, the findings imply that crisis management should focus on reducing negative emotions by highlighting shared responsibility. For regulators, the study underscores the need to address the moral and emotional dimensions of public opinion to facilitate successful risk communication and technology adoption.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• acceptance adoption

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: self report data