{
  "fetched_at": "2026-10-20T01:11:46Z",
  "source_url": "https://www.kaggle.com/competitions/2022-spring-competition-psych-5710/data",
  "pages": [
    {
      "name": "data-description",
      "content": "<h2>File descriptions</h2>\n<ul>\n<li><strong>full.data.Rdata</strong> - the training and testing sets</li>\n<li><strong>sampleSubmission.csv</strong> - a sample submission file in the correct format</li>\n\n<h2>Data Set Information:</h2>\n\n<p> The dataset provides patient reviews on specific drugs along with related conditions and a 10 star patient rating reflecting overall patient satisfaction. The data was obtained by crawling online pharmaceutical review sites. The original data was downloaded from the UCI Machine Learning repository here:</p>\n\n<a href=\"https://archive.ics.uci.edu/ml/datasets/Drug+Review+Dataset+%28Drugs.com%29\"> Original Drug Review Dataset</a>\n\n<h2>Important notes:</h2>\n\n<p> When using this dataset, you agree that you\n\n1) only use the data for research purposes\n\n2) don't use the data for any commercial purposes\n\n3) don't distribute the data to anyone else\n\n4) cite the authors.</p>\n\n\n<h2>Attribute Information:</h2>\n\n<p>doc_id: ID of the patient</p>\n\n<p> text: patient review</p>\n\n<p>rating (numerical): the Z-score of the 10 star patient rating</p>\n\n<p>Weekday: what was the day of the week when the review was posted</p>\n\n<p>Month: what was the month when the review was posted</p>\n\n<p>Year: what was the year when the review was posted</p>\n\n<p>usefulCount (numerical): the Z-score of the number of users who found review useful</p>\n\n<p> DataGroup: Does the review belongs to the \"Training\" or to the \"Testing\" sets? You'll need to train your models in the \"Training\" set, and generate predictions of \"rating\" for the \"Testing\" set. </p>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "<h2>Introduction</h2>\n\nHi, it's time to apply everything you learned in our Introduction to Machine Learning and Data Mining course.\n\n<p>In this competition you'll work with a dataset containing drug reviews (i.e. textual information that people that used some medications posted in an online medical portal). The dataset provides patient reviews on specific drugs along with related conditions and a 10 star patient rating reflecting overall patient satisfaction. Your goal is to predict the patient satisfaction score.</p>\n\nWe'll have two leaderboards: Public and Private.\n\nThe public leaderboard will calculate the RMSE based on 50% of the test set, while the private leaderboard will be based on the remaining 50%.\n\nYou can make up to 10 submissions per day. Participants will need to wait until the next UTC day after submitting the maximum number of daily submissions.\n\nAlso, you can select up to 2 submissions as the number of submissions eligible for the final private leaderboard. Users can hand-select the eligible submissions, or will otherwise default to the best public scoring submissions.\n\nGood luck, and have fun!\n\n\n<h2>Grading</h2>\n\n<p>**The full grade for this competition will be worth 40 points**. Your final grade will depend on your** final RMSE**, i.e. in the RMSE **calculated using the complete test set (average of the RMSE obtained in the Public and in the Private Leaderboard)**. Notice that once you submit your predictions to the Kaggle platform, the resulting RMSE will be calculated using only half of the test set. So, make sure that you are paying attention to weather or not your predictive model is stable (i.e. presents low variance) when applied to new, unseen datasets.</p>\n\n<p>Your grade will depend on your performance. You will receive a **full grade** (100%) if your final **RMSE is equal to or below 0.8795**. </p>\n\n<p> If your final **RMSE is between 0.8795 and 0.8825** you'll receive **90% of the total grade** for the competition.</p>\n\n<p> If your final **RMSE is between 0.8826 and 0.8910** you'll receive **70% of the total grade** for the competition.</p>\n\n<p> If your final **RMSE is between 0.8911 and 0.8964** you'll receive **50% of the total grade** for the competition.</p>\n\n<p> If your final **RMSE is between 0.8965 and 0.901** you'll receive **25% of the total grade** for the competition.</p>\n\n<p> Finally, if your RMSE is above 0.902 you'll receive 0% of the total grade for the competition</p>\n\n<h2>Data Set Information:</h2>\n\n<p> The dataset provides patient reviews on specific drugs along with related conditions and a 10 star patient rating reflecting overall patient satisfaction. The data was obtained by crawling online pharmaceutical review sites. The original data was downloaded from the UCI Machine Learning repository here:</p>\n\n<a href=\"https://archive.ics.uci.edu/ml/datasets/Drug+Review+Dataset+%28Drugs.com%29\"> Original Drug Review Dataset</a>\n\n<h2>Important notes:</h2>\n\n<p> When using this dataset, you agree that you\n\n1) only use the data for research purposes\n\n2) don't use the data for any commercial purposes\n\n3) don't distribute the data to anyone else\n\n4) cite the authors.</p>\n\n\n<h2>Attribute Information:</h2>\n\n<p>doc_id: ID of the patient</p>\n\n<p> text: patient review</p>\n\n<p>rating (numerical): the Z-score of the 10 star patient rating</p>\n\n<p>Weekday: what was the day of the week when the review was posted</p>\n\n<p>Month: what was the month when the review was posted</p>\n\n<p>Year: what was the year when the review was posted</p>\n\n<p>usefulCount (numerical): the Z-score of the number of users who found review useful</p>\n\n<p> DataGroup: Does the review belongs to the \"Training\" or to the \"Testing\" sets? You'll need to train your models in the \"Training\" set, and generate predictions of \"rating\" for the \"Testing\" set. </p>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "Grading\n\n**The full grade for this competition will be worth 40 points**. Your final grade will depend on your** final RMSE**, i.e. in the RMSE **calculated using the complete test set (average of the RMSE obtained in the Public and in the Private Leaderboard)**. Notice that once you submit your predictions to the Kaggle platform, the resulting RMSE will be calculated using only half of the test set. So, make sure that you are paying attention to weather or not your predictive model is stable (i.e. presents low variance) when applied to new, unseen datasets.\n\nYour grade will depend on your performance. You will receive a **full grade** (100%) if your final **RMSE is equal to or below 0.8795**.\n\nIf your final **RMSE is between 0.8795 and 0.8825** you'll receive **90% of the total grade** for the competition.\n\nIf your final **RMSE is between 0.8826 and 0.8910** you'll receive **70% of the total grade** for the competition.\n\nIf your final **RMSE is between 0.8911 and 0.8964** you'll receive **50% of the total grade** for the competition.\n\nIf your final **RMSE is between 0.8965 and 0.901** you'll receive **25% of the total grade** for the competition.\n\nFinally, if your RMSE is above 0.902 you'll receive 0% of the total grade for the competition\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": " - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}