Back to Search Start Over

Aggregating Google Trends: Multivariate Testing and Analysis

Authors :
France, Stephen L.
Shi, Yuying
Publication Year :
2017

Abstract

Web search data are a valuable source of business and economic information. Previous studies have utilized Google Trends web search data for economic forecasting. We expand this work by providing algorithms to combine and aggregate search volume data, so that the resulting data is both consistent over time and consistent between data series. We give a brand equity example, where Google Trends is used to analyze shopping data for 100 top ranked brands and these data are used to nowcast economic variables. We describe the importance of out of sample prediction and show how principal component analysis (PCA) can be used to improve the signal to noise ratio and prevent overfitting in nowcasting models. We give a finance example, where exploratory data analysis and classification is used to analyze the relationship between Google Trends searches and stock prices.

Subjects

Subjects :
Economics - Econometrics

Details

Database :
arXiv
Publication Type :
Report
Accession number :
edsarx.1712.03152
Document Type :
Working Paper