This query is specific to the file "train-item-views.csv".
The timeframe column is a bit misleading. The time column is NOT a timestamp which makes it difficult to figure out what is the true sequence if items within a session. I am assuming that the timeframe column tells us the time elapsed (in miliseconds) since the user first entered a search query on the retail platform till the moment the item was clicked.
Now my question is, should we sort each session in non-descending order of the timeframe? To get the correct sequence of items in a session? If yes then why isn't the datafile already sorted. If NOT, then what is meaning or use of the timeframe column?
Posted by: abanerjee @ Oct. 20, 2021, 7:16 a.m.when I find the diginetica dataset
Posted by: nourelhouda.aboub @ June 7, 2022, 10:42 p.m.