Guns as Symbols in Ads

by

Faculty Sponsor: Erika Franklin Fowler

Live Poster Session: Zoom Link

Latonya Smith
Latonya Smith

Latonya Smith is a rising junior(‘24). She hails from Jamaica and has decided to pursue Computer Science and Economics as a major here at Wesleyan. She is involved in the Wesleyan Boxing club and is an Eco-Facilitator. Her career interests are Data Science, Machine Learning and Software Engineering.

Ignore this button 🙂
Emma Tuhabonye
Emma Tuhabonye

Emma Tuhabonye is a rising senior (’23)  from Austin, Texas. She graduated high school from Saint Andrew’s Episcopal School. Her interests include Solitaire, running, swimming, and machine learning. Emma is involved the Wesleyan Track and Field team and is a member of Wesleyan Women in Science (WesWIS). She is majoring in Computer Science with a minor and Data Analysis. After graduation, Emma hopes to pursue a career as a Data Scientist.

Ignore this button 🙂

Abstract: Recent increases in mass shootings have resulted in more political discussions about guns and subsequently gun laws. As a result, researchers have focused on analyzing political ads as a way to examine the discourse in reference to these issues. Our data set was the TV advertisements sponsored by federal candidates for the 2020 general election. This project focuses on the detection of gun images and the mention of guns in political television advertisements. We used and compared deep learning techniques such as object detection and image classification to analyze the images and videos in these ads and used logistic regression to determine the factors that influence the decision to mention guns or use their images. We also used Fightin Words(Monroe et al 2008) to assess the distinct terms that are used by Democrats and Republicans in TV political ads that mention guns. Our findings revealed that party affiliation only matters for gun mentionings. We also saw that Democrats and Republicans talk about gun issues in different ways. From this we decided on our next steps to further our research.

GUN_QAC_07252022-3