Publicação

Impact of gun violence on restaurants: analysis of vader-derived sentiment in yelp reviews

Ver documento

Detalhes bibliográficos
Resumo:This paper examines how mass shootings influence consumer perceptions of restaurants using over 900,000 Yelp reviews from 50,000 U.S. restaurants. Staggered Difference-in-Differences (DiD) and sentiment analysis (VADER, BERT) are employed to assess changes in sentiment, review length, ratings, and frequency at distances of 5 km, 10 km, and 20 km post-incident. Contrary to expectations, limited evidence is found that mass shootings significantly affect sentiment, ratings, or review volume. Although minor shifts occur—fewer food-oriented comments and occasional mentions of shootings—no consistent negative patterns emerge. These findings suggest subtler or undetected impacts not fully captured by online reviews.
Autores principais:Otott-Kovács, Gábor
Assunto:Gun violence Yelp Restaurants Review Text sentiment Big data Difference-in differences Staggered difference-in-differences Consumer behavior Online word of mouth Gun violence archive NLP Causal inference VADER Business analytics Business and data analytics Data Data visualization LLM Mistral BERT
Ano:2025
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso aberto
Instituição associada:Universidade Nova de Lisboa
Idioma:inglês
Origem:Repositório Institucional da UNL
Descrição
Resumo:This paper examines how mass shootings influence consumer perceptions of restaurants using over 900,000 Yelp reviews from 50,000 U.S. restaurants. Staggered Difference-in-Differences (DiD) and sentiment analysis (VADER, BERT) are employed to assess changes in sentiment, review length, ratings, and frequency at distances of 5 km, 10 km, and 20 km post-incident. Contrary to expectations, limited evidence is found that mass shootings significantly affect sentiment, ratings, or review volume. Although minor shifts occur—fewer food-oriented comments and occasional mentions of shootings—no consistent negative patterns emerge. These findings suggest subtler or undetected impacts not fully captured by online reviews.