Analysis of the Impact of the US Presidential Election on the US Economy Based on Natural Language Processing and Big Data

Abstract

Based on the background of the 2020 US presidential election, this paper analyzes the impact of the two candidates’ different policy propositions on the US economy. Firstly, using web crawler technology to obtain the public information, then using Natural Language Processing to generate the theme of the document and corresponding high frequency keywords. Secondly, build a mathematical model. GDP, CPI and unemployment are selected as the macro indicator system to measure economic changes. In the stage of data collection and preprocessing, which also spent a lot of time to ensure that the quality of data can be applied to the establishment of subsequent models. In the model part, this paper uses the multiple linear regression model. Based on this, it objectively gives significant indicators to measure the size and direction of the influence of the two candidates’ positions on the American economy. Finally, based on the time series data, this paper uses the ARIMA Model to predict the next four periods, and gives the economic trend of the United States in the next year if President Trump continues to be in office.