{
  "id": 253087,
  "title": "relations between user_click_history/exposed_items & labels/time",
  "url": "/competitions/bigdata2021-rl-recsys/discussion/253087",
  "author_name": "",
  "post_date": "2021-07-15T00:23:58.785984900Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hello!</p>\n<p>We're trying to understand a little better how users interacts with this item store and what the datasets represent on a typical user session. </p>\n<p>To our understanding exposed_items is a set of items that can be purchased by a user. A user can only purchase one item at a time and only 3 items are available for purchase. Once all three items are bought, the next set of items be available for sale. </p>\n<ol>\n<li><p>Would the user_click_history represent a prior run of the game? (so they've bought these items before)</p></li>\n<li><p>Is the click history also a user_purchase_history? If not, what does user_click_history represent? </p></li>\n<li><p>Are all items in the store available to be viewed from the beginning (and the user just knows the order of the following 9)?</p></li>\n<li><p>The dataset also includes a timestamp feature, can we assume this timestamp was the last purchase item time? Would that also mean we can assume there were prior versions of the label feature with older timestamps?</p></li>\n</ol>\n<p>Thanks!</p>",
  "messages": [
    {
      "id": "1388437",
      "postDate": "07/15/2021 00:23:58",
      "content": "<p>Hello!</p>\n<p>We're trying to understand a little better how users interacts with this item store and what the datasets represent on a typical user session. </p>\n<p>To our understanding exposed_items is a set of items that can be purchased by a user. A user can only purchase one item at a time and only 3 items are available for purchase. Once all three items are bought, the next set of items be available for sale. </p>\n<ol>\n<li><p>Would the user_click_history represent a prior run of the game? (so they've bought these items before)</p></li>\n<li><p>Is the click history also a user_purchase_history? If not, what does user_click_history represent? </p></li>\n<li><p>Are all items in the store available to be viewed from the beginning (and the user just knows the order of the following 9)?</p></li>\n<li><p>The dataset also includes a timestamp feature, can we assume this timestamp was the last purchase item time? Would that also mean we can assume there were prior versions of the label feature with older timestamps?</p></li>\n</ol>\n<p>Thanks!</p>",
      "rawMarkdown": "Hello!\n\nWe're trying to understand a little better how users interacts with this item store and what the datasets represent on a typical user session. \n\nTo our understanding exposed_items is a set of items that can be purchased by a user. A user can only purchase one item at a time and only 3 items are available for purchase. Once all three items are bought, the next set of items be available for sale. \n\n1. Would the user_click_history represent a prior run of the game? (so they've bought these items before)\n\n2. Is the click history also a user_purchase_history? If not, what does user_click_history represent? \n\n3. Are all items in the store available to be viewed from the beginning (and the user just knows the order of the following 9)?\n\n4. The dataset also includes a timestamp feature, can we assume this timestamp was the last purchase item time? Would that also mean we can assume there were prior versions of the label feature with older timestamps?\n\nThanks!",
      "votes": null
    },
    {
      "id": "1388456",
      "postDate": "07/15/2021 01:23:28",
      "content": "<p>So if you need all the attributes that covers users interactions on the web, I will highly recommend you to have a look at the various fields covered by Google Analytics. A sample data can be viewed here <a href=\"https://www.kaggle.com/bigquery/google-analytics-sample?select=ga_sessions_20160813\" target=\"_blank\">https://www.kaggle.com/bigquery/google-analytics-sample?select=ga_sessions_20160813</a></p>\n<p>Now to get all the user clicks or capture what a user did you might have to analyze the database as well which might represent the various activities of a user.</p>\n<p><strong><em>Do Upvote if the answer was helpful :)</em></strong></p>",
      "rawMarkdown": "So if you need all the attributes that covers users interactions on the web, I will highly recommend you to have a look at the various fields covered by Google Analytics. A sample data can be viewed here [https://www.kaggle.com/bigquery/google-analytics-sample?select=ga_sessions_20160813](https://www.kaggle.com/bigquery/google-analytics-sample?select=ga_sessions_20160813)\n\nNow to get all the user clicks or capture what a user did you might have to analyze the database as well which might represent the various activities of a user.\n\n***Do Upvote if the answer was helpful :)***",
      "votes": null
    },
    {
      "id": "1388952",
      "postDate": "07/15/2021 10:38:53",
      "content": "<p>I can't answer question 1, 2, and 4 considering the anonymity of the dataset. You can use the provided features in any way. <br>\nFor Q3, all nine items are visible to users at the beginning.</p>",
      "rawMarkdown": "I can't answer question 1, 2, and 4 considering the anonymity of the dataset. You can use the provided features in any way. \nFor Q3, all nine items are visible to users at the beginning.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1388456,
      "author_name": "shivkumarganesh",
      "author_url": "",
      "post_date": "07/15/2021 01:23:28",
      "content": "<p>So if you need all the attributes that covers users interactions on the web, I will highly recommend you to have a look at the various fields covered by Google Analytics. A sample data can be viewed here <a href=\"https://www.kaggle.com/bigquery/google-analytics-sample?select=ga_sessions_20160813\" target=\"_blank\">https://www.kaggle.com/bigquery/google-analytics-sample?select=ga_sessions_20160813</a></p>\n<p>Now to get all the user clicks or capture what a user did you might have to analyze the database as well which might represent the various activities of a user.</p>\n<p><strong><em>Do Upvote if the answer was helpful :)</em></strong></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1388952,
      "author_name": "sword10086",
      "author_url": "",
      "post_date": "07/15/2021 10:38:53",
      "content": "<p>I can't answer question 1, 2, and 4 considering the anonymity of the dataset. You can use the provided features in any way. <br>\nFor Q3, all nine items are visible to users at the beginning.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1388437": "Hello!\n\nWe're trying to understand a little better how users interacts with this item store and what the datasets represent on a typical user session. \n\nTo our understanding exposed_items is a set of items that can be purchased by a user. A user can only purchase one item at a time and only 3 items are available for purchase. Once all three items are bought, the next set of items be available for sale. \n\n1. Would the user_click_history represent a prior run of the game? (so they've bought these items before)\n\n2. Is the click history also a user_purchase_history? If not, what does user_click_history represent? \n\n3. Are all items in the store available to be viewed from the beginning (and the user just knows the order of the following 9)?\n\n4. The dataset also includes a timestamp feature, can we assume this timestamp was the last purchase item time? Would that also mean we can assume there were prior versions of the label feature with older timestamps?\n\nThanks!",
    "1388456": "So if you need all the attributes that covers users interactions on the web, I will highly recommend you to have a look at the various fields covered by Google Analytics. A sample data can be viewed here [https://www.kaggle.com/bigquery/google-analytics-sample?select=ga_sessions_20160813](https://www.kaggle.com/bigquery/google-analytics-sample?select=ga_sessions_20160813)\n\nNow to get all the user clicks or capture what a user did you might have to analyze the database as well which might represent the various activities of a user.\n\n***Do Upvote if the answer was helpful :)***",
    "1388952": "I can't answer question 1, 2, and 4 considering the anonymity of the dataset. You can use the provided features in any way. \nFor Q3, all nine items are visible to users at the beginning."
  },
  "source": "meta"
}