Prediction Service - R&D

Prediction Service/R&D/datasets/wikipedia/1_download_data.sh

+# store description
+curl http://www.wikibench.eu/wiki/2007-09/ > wikibench.repo
+
+# download each file
+cat wikibench.repo | grep -oh "wiki\.[0-9]\+\.gz" | xargs -I{} curl -O http://www.wikibench.eu/wiki/2007-09/{}
+

Prediction Service/R&D/datasets/wikipedia/2_get_subset.sh

+for file in raw/*
+do
+    echo $file
+    gunzip -c $file > ./temp
+
+    grep de.wikipedia.org ./temp >> result.txt
+done
\ No newline at end of file

Prediction Service/R&D/datasets/wikipedia/3_filter_timestamps.sh

+# extract only the timestamps
+cat result.txt | awk '{print $2}' | grep -oh "^[0-9]\+" > requests.txt

Prediction Service/README.md

+# Prediction Service
+
+- In the **R&D folder**, you can find how to explore data and develop prediction algorithms. You can also find the validation of the algorithm against Wikipedia JP and DE.
+
+*The idea of prediction*
+![main and compared vectors.png](main and compared vectors.png)
+
+*Results*
+![Results.png](Results.png)
+
+- The **Application folder** contains the Prediction Service (Python APP)