Refractive datasets as a sensemaking methodology in closed data ecosystems

Data, code, and supplement for:

Beers, A., Ito, V., Orozco, A., Gildersleve, P., Aragón, P., & Tripodi, F. (2025). Refractive datasets as a sensemaking methodology in closed data ecosystems. Big Data & Society, 12(4). https://doi.org/10.1177/20539517251406193 (Original work published 2025).

The following files are uploaded:

1. RefractiveDatasets_Code.py. This is commented Python code for reproducing analyses in this paper. It assumes that any data downloaded will be present in a folder titled Data located one level up from where you run the script ("../Data"). Some data can be regenerated via this code; other data is provided with an explanation of how to independently retrieve it in the comments.

2. Figure2Data.csv. Time series data for Figure 2, representing Wikipedia traffic across different Wikipedia projects.

3. Figure3Data.csv. Time series data for Figure 4, representing the median of five data pulls of Google Trends traffic for a given keyword.

4. GoogleTrends_DataSamples.zip. Zipped CSV files corresponding to five unique data downloads from Google Trends used to construct Figure3Data.csv.

5. GoogleTrendsAnalysisSupplement.pdf. A supplemental analysis of the robustness of Google Trends data portrayed in Figure 3.

6. Figure45_HurricaneClusterArticles.csv. A list of articles found in the "Atlantic Hurricanes" cluster identified in Figures 4 and 5.

7. Figure45_ClusteredData.zip. A compressed GEXF network file representing clustered correlation relationships between articles portrayed in Figures 4 and 5.

8. Figure45.gephi. A Gephi visualization file for the network visualization portrayed in Figures 4 and 5.