Collaborative Research: Understanding the Evolution of Political Campaign Advertisements over the Last Century

电视广告是选民在没有媒体过滤的情况下了解候选人的主要方式，在美国的政治竞选中无处不在。本项目利用现有但未充分利用的超过10万份政治广告数据集，审查政治广告的演变，特别是在问题宣传和协调方面。该项目为政治学、信息学和图书馆学等多个领域做出了重要的方法论和实质性贡献，并增强了我们对广告塑造和告知政治行为方式的理解。它还促进研究和教学机构在新兴的自动视听分析子领域的跨学科研究生和本科教育，包括来自传统上代表性不足的群体的学生和学者。该项目使用了从1912年到2018年超过10万个政治广告的集合，即俄克拉何荷马大学存档的朱利安·p·坎特政治商业集，以开发(1)一个自动化系统，从广告图像、音频和文本中识别问题（和其他政治相关）内容；(2)一个最先进的用户界面，使研究人员能够查询、交互和查看视频，并输出数据进行分析。该项目使用数据来检查政治广告的演变，测试问题的横向和纵向扩散模型。该项目的工具和数据是在一个跨学科研讨会上介绍的，可供研究美国政治、竞选、政治传播和公众舆论的多个学科的学者广泛使用。该合作项目由问责机构和行为计划（AIB）和促进竞争研究的既定计划（EPSCoR）共同资助。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Television advertising, a primary way voters hear about candidates absent a media filter, is ubiquitous with political campaigns in the United States. Drawing on an existing but underutilized data set of over 100,000 political ads, this project examines the evolution of political advertising, especially as it pertains to issue advocacy and consonance. The project makes significant methodological and substantive contributions to several fields including political, information, and library sciences, and enhances our understanding of the way ads shape and inform political behavior. It also promotes interdisciplinary graduate and undergraduate education at research and teaching institutions in the emerging sub-field of automated audiovisual analysis, including students and scholars from traditionally underrepresented groups.The project uses a collection of over 100,000 political ads from 1912-2018, the Julian P. Kanter Political Commercial Collection archived at the University of Oklahoma, to develop (1) an automated system to identify issue (and other politically relevant) content from ad image, audio, and text and (2) a state-of-the-art user interface that gives researchers the ability to query, interact with and view videos, and also output data for analysis. The project uses the data to examine the evolution of political advertising, testing models of horizontal and vertical diffusion of issues. The project’s tools and data, introduced at an interdisciplinary workshop, are widely available to scholars across a number of disciplines who study American politics, campaigns, political communication and public opinion.This collaborative project is jointly funded by the Accountable Institutions and Behavior (AIB) program and the Established Program to Stimulate Competitive Research (EPSCoR).This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.