{
  "id": 364260,
  "title": "Some business specific questions about the products(aids) and website.",
  "url": "/competitions/otto-recommender-system/discussion/364260",
  "author_name": "Tarik Karakas",
  "post_date": "2022-11-05T13:33:51.285000",
  "votes": 15,
  "comment_count": 0,
  "views": 0,
  "content": "<ol>\n<li><p>We have <code>1855603</code> unique products in training data. We don't know anything about products and their categories or sellers. At least we have to know category ids to recommend related products, otherwise we have to cluster these products, and it seems like meanless effort for me. Do you planning to provide this information?</p></li>\n<li><p>OTTO has <code>related products</code> recommendation in the website. For example: If a customer clicks on any product (<a href=\"https://www.otto.de/p/samsung-galaxy-a52s-5g-smartphone-16-4-cm-6-5-zoll-128-gb-speicherplatz-64-mp-kamera-1481253571/#variationId=1481253572\" target=\"_blank\">Samsung a52s 128gb</a>), they are more likely to click on the recommended product (in this case: <a href=\"https://www.otto.de/p/samsung-galaxy-a52s-smartphone-16-4-cm-6-5-zoll-256-gb-speicherplatz-64-mp-kamera-1481274370/#variationId=1481274379\" target=\"_blank\">Samsung a52s 256gb</a>). We must know these recommended products to predict the next item. </p></li>\n</ol>\n<p>It is just an idea to get more robust results. Thanks for your answer in advance.</p>",
  "messages": [
    {
      "id": 2018196,
      "postDate": "2022-11-05T13:33:51.287Z",
      "content": "<ol>\n<li><p>We have <code>1855603</code> unique products in training data. We don't know anything about products and their categories or sellers. At least we have to know category ids to recommend related products, otherwise we have to cluster these products, and it seems like meanless effort for me. Do you planning to provide this information?</p></li>\n<li><p>OTTO has <code>related products</code> recommendation in the website. For example: If a customer clicks on any product (<a href=\"https://www.otto.de/p/samsung-galaxy-a52s-5g-smartphone-16-4-cm-6-5-zoll-128-gb-speicherplatz-64-mp-kamera-1481253571/#variationId=1481253572\" target=\"_blank\">Samsung a52s 128gb</a>), they are more likely to click on the recommended product (in this case: <a href=\"https://www.otto.de/p/samsung-galaxy-a52s-smartphone-16-4-cm-6-5-zoll-256-gb-speicherplatz-64-mp-kamera-1481274370/#variationId=1481274379\" target=\"_blank\">Samsung a52s 256gb</a>). We must know these recommended products to predict the next item. </p></li>\n</ol>\n<p>It is just an idea to get more robust results. Thanks for your answer in advance.</p>",
      "rawMarkdown": "1. We have `1855603` unique products in training data. We don't know anything about products and their categories or sellers. At least we have to know category ids to recommend related products, otherwise we have to cluster these products, and it seems like meanless effort for me. Do you planning to provide this information?\n\n2.  OTTO has `related products` recommendation in the website. For example: If a customer clicks on any product ([Samsung a52s 128gb](https://www.otto.de/p/samsung-galaxy-a52s-5g-smartphone-16-4-cm-6-5-zoll-128-gb-speicherplatz-64-mp-kamera-1481253571/#variationId=1481253572)), they are more likely to click on the recommended product (in this case: [Samsung a52s 256gb](https://www.otto.de/p/samsung-galaxy-a52s-smartphone-16-4-cm-6-5-zoll-256-gb-speicherplatz-64-mp-kamera-1481274370/#variationId=1481274379)). We must know these recommended products to predict the next item. \n\nIt is just an idea to get more robust results. Thanks for your answer in advance.",
      "votes": 15
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2018196": "1. We have `1855603` unique products in training data. We don't know anything about products and their categories or sellers. At least we have to know category ids to recommend related products, otherwise we have to cluster these products, and it seems like meanless effort for me. Do you planning to provide this information?\n\n2.  OTTO has `related products` recommendation in the website. For example: If a customer clicks on any product ([Samsung a52s 128gb](https://www.otto.de/p/samsung-galaxy-a52s-5g-smartphone-16-4-cm-6-5-zoll-128-gb-speicherplatz-64-mp-kamera-1481253571/#variationId=1481253572)), they are more likely to click on the recommended product (in this case: [Samsung a52s 256gb](https://www.otto.de/p/samsung-galaxy-a52s-smartphone-16-4-cm-6-5-zoll-256-gb-speicherplatz-64-mp-kamera-1481274370/#variationId=1481274379)). We must know these recommended products to predict the next item. \n\nIt is just an idea to get more robust results. Thanks for your answer in advance."
  }
}