{
  "fetched_at": "2026-10-20T09:33:31Z",
  "source_url": "https://www.kaggle.com/competitions/dlhlp2020spring-asr/data",
  "pages": [
    {
      "name": "rules",
      "content": "This is a page where you can include rules that participants must accept before joining. You may wish to include rules like:\n\n - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "data-description",
      "content": "<p>This page shows up underneath the data files. </p>\n\n## Get Data\n- **Data Download Link:** https://drive.google.com/file/d/1daFU8tPPUyhN7Fc6JUTohEfHXIn6ZDgq\n- extraction: `tar zxvf DLHLP.tar.gz`\n  if you encounter errors like `Cannot change ownership to uid`, please add `--no-same-owner`.\n\n## Data fields\n\n### train/\n  - audio files from `000001.wav` to `008000.wav`.\n  - transcription text `train/bopomo.trans.txt`\n    - format: each line contains `<index><space><ㄅㄆㄇㄈ transcription>`\n\n### dev/\n  - audio files from `008001.wav` to `009000.wav`.\n  - transcription text `dev/bopomo.trans.txt`\n    - format: each line contains `<index><space><ㄅㄆㄇㄈ transcription>`\n\n### test/\n  - audio files from `009001.wav` to `010000.wav`.\n  - transcription text `test/bopomo.trans.txt` is still attached for consistency, but all transcription are replaced with fake transcription.\n    - format: each line contains `<index><space><ㄅㄆㄇㄈ fake transcription>`\n\n### sample_submission.txt\n  - <strong>id</strong> - an audio file id. \n  - <strong>answer</strong> - the bopomofo sequence of the according wav audio file.\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "Welcome to DLHLP2020 Homework 1, end-to-end speech recognition for Bopomofo characters. \n\n## Homework description\n[Slide link](https://docs.google.com/presentation/d/1C2a1_V91EHjy3sgxy3rKWqU6hCIOOsbMWuOhYCZW5jQ/edit?usp=sharing)\n\n## Kaggle Upload Deadline\n2020/03/22 (Sun.) 23:59 (UTC+8)\n\n## Evaluation Metric\n\n**Mean [Levenshtein Distance](https://en.wikipedia.org/wiki/Levenshtein_distance)** calculated for each sentence in char-level.\n\nLevenshtein distance between two character sequence (predicted sequence v.s. target sequence) is the minimum number of single-character edits (insertions, deletions or substitutions) required to change one word into the other.\n\n<p>\\[ LevDistance = N_{ins}+N_{del}+N_{sub} \\]</p>\n\np.s. **Word-Error Rate (WER)** is normalized Levenshtein Distance (to target sequence length), a more common evaluation metric for ASR.\n\n<p>\\[ WER = \\frac{N_{ins}+N_{del}+N_{sub}}{N_{target-length}} \\]</p>\n\n<h2>Submission Format</h2>\n<p><strong>All student should submit your answer in CSV format with <code>,</code> as delimiter</strong>. Submission files should contain two columns: <code>id</code> and <code>answer</code>. <code>id</code> should be a six-digits number (e.g. <code>009001</code> according to <code>009001.wav</code> audio file. The answer to the wav file should be put in <code>answer</code> column. There should <strong>NOT</strong> be any space between the <code>answer</code> and delimiter. Sorting according to `id` is not necessary.\n\n<p>There is an example format:</p>\n<pre>id,answer<br /><span style=\"font-size: 14px; line-height: 1.4em;\">009001,ㄩˇ ㄧㄣ ㄅㄧㄢˋ ㄕˋ<br />009002,ㄩˇ ㄧㄣ ㄅㄧㄢˋ ㄕˋ<br /></span><span style=\"font-size: 14px; line-height: 1.4em;\">...,...<br /></span></span><span style=\"font-size: 14px; line-height: 1.4em;\">009999,ㄩˇ ㄧㄣ ㄅㄧㄢˋ ㄕˋ<br /></span><span style=\"font-size: 14px; line-height: 1.4em;\">010000,ㄩˇ ㄧㄣ ㄅㄧㄢˋ ㄕˋ</span></pre>\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "<p>The evaluation pages describes how submissions will be scored and how students should format their submissions. </p>\n<hr />\n\n## Evaluation Metric\n\nThe evaluation metric for this competition is **Mean [Levenshtein Distance](https://en.wikipedia.org/wiki/Levenshtein_distance)** calculated for each sentence in char-level.\n\nLevenshtein distance between two character sequence (predicted sequence v.s. target sequence) is the minimum number of single-character edits (insertions, deletions or substitutions) required to change one word into the other.\n\n<p>\\[ LevDistance = N_{ins}+N_{del}+N_{sub} \\]</p>\n\n**Word-Error Rate (WER)** is normalized Levenshtein Distance (to target sequence length), a more common evaluation metric for ASR.\n\n<p>\\[ WER = \\frac{N_{ins}+N_{del}+N_{sub}}{N_{target-length}} \\]</p>\n\n<h2>Submission Format</h2>\n<p><strong>All student should submit your answer in CSV format with <code>,</code> as delimiter</strong>. Submission files should contain two columns: <code>id</code> and <code>answer</code>. <code>id</code> should be a six-digits number (e.g. <code>009001</code> according to <code>009001.wav</code> audio file. The answer to the wav file should be put in <code>answer</code> column. There should <strong>NOT</strong> be any space between the <code>answer</code> and delimiter. Sorting according to `id` is not necessary.\n\n<p>There is an example format:</p>\n<pre>id,answer<br /><span style=\"font-size: 14px; line-height: 1.4em;\">009001,ㄩˇ ㄧㄣ ㄅㄧㄢˋ ㄕˋ<br />009002,ㄩˇ ㄧㄣ ㄅㄧㄢˋ ㄕˋ<br /></span><span style=\"font-size: 14px; line-height: 1.4em;\">...,...<br /></span></span><span style=\"font-size: 14px; line-height: 1.4em;\">009999,ㄩˇ ㄧㄣ ㄅㄧㄢˋ ㄕˋ<br /></span><span style=\"font-size: 14px; line-height: 1.4em;\">010000,ㄩˇ ㄧㄣ ㄅㄧㄢˋ ㄕˋ</span></pre>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}