RAPID: Election Result Anomaly Detection for 2020

在我们这个时代，金钱和思想在互联网上迅速传播，“所有的政治都是地方的”这句老话可能不像以前那么正确了。但美国的选举管理仍然顽固地保持着地方性。因此，选举结果由州和地方机构以特殊格式公布。任何对全国选举结果的研究，甚至是跨几个州的选举结果的研究，都需要费力地收集不同的格式。这个项目建立了一个系统，可以有效地将来自多个州的选举结果数据集汇编成单一格式。此外，该项目还实现了各种算法，用于快速、广泛地检测和可视化选举结果中的异常。该系统将迅速标记2020年11月大选的初步选举结果数据中的重大异常。这项分析将公开提供给候选人、政党和选举行政人员。候选人、政党和选举管理人员对当地情况有详细的了解，可以评估他们所在地区的任何异常现象是否有合理的解释，或者是否需要进一步调查。在形势最好的时候，进行一场值得信赖的选举也会带来许多后勤方面的挑战。2020年，除了最近选举技术的变化和广泛承认的网络攻击威胁外，还有一个前所未有的复杂因素，即在总统选举年，一场肆虐的流行病。该项目将为选举核查工具箱提供一个重要的工具，帮助在选举结果最终得到认证之前识别异常情况。候选人和其他利益相关方将能够利用这一分析及时做出决定，决定是质疑还是要求对选举结果进行调查——在确定获胜者之前。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。

In our era, with money and ideas traveling quickly over the internet, the old saw that “all politics is local” may be less true than it used to be. But the administration of elections in the United States remains doggedly local. As a result, election results are published by state and local agencies in idiosyncratic formats. Any study of election results at a national level, or even across several states, requires painstaking assembly of disparate formats. This project builds a system to efficiently assemble election result data sets from several states into a single format. In addition, the project implements a variety of algorithms for fast, broad detection and visualization of anomalies in election results. The system will quickly flag significant anomalies in preliminary election result data from the general election in November 2020. This analysis will be made available publicly and to candidates, parties and election administrators. Candidates, parties and election administrators have the detailed local knowledge required to assess whether any particular anomaly in their district has a legitimate explanation or whether further investigation is appropriate.In the best of times, conducting a trustworthy election poses many logistical challenges. In 2020, in addition to recent changes in election technology and widely acknowledged threats of cyberattack, there is the unprecedented complication of a raging pandemic in a Presidential election year. The project will provide a significant tool in the election verification toolbox by helping to identify anomalies well before election results are finally certified. Candidates and other interested parties will be able to use this analysis to inform timely decisions about whether to challenge or demand investigations into election results – before the determination of winners is set in stone.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.