Skip to content

Latest commit

 

History

History
16 lines (11 loc) · 1.45 KB

README.md

File metadata and controls

16 lines (11 loc) · 1.45 KB
Raw

Calibration of Google Trends Time Series

This directory contains all code and data to reproduce the results published in this paper:

Robert West. Calibration of Google Trends Time Series. In Proceedings of the 29th ACM International Conference on Information and Knowledge Management (CIKM). 2020. [PDF]

Note that, whereas the results in the paper were produced with an R implementation, we have since switched over to Python. The R implementation that was used for the paper and that is contained in this directory is not updated anymore. If you are interested in using Google Trends Anchor Bank (G-TAB) in your own research, please use the Python implementation instead.

Directory contents:

  • The master script for reproducing the paper is src/R/code_for_reproducing_cikm2020_paper.Rmd.
  • An evaluated, HTML-rendered version of the script is available here.
  • Data can be found in data/; in particular, data/calibration contains the cached results of Google Trends requests for constructing the anchor bank in the offline preprocessing stage (Sec. 2.1 of the paper); data/binsearch contains cached results of Google Trends requests from the online binary search phase (Sec. 2.2 of the paper).
  • All plots from the paper can be found in plots/.