YouTube-Based TOEFL Learning Transcript Dataset
Description
The YouTube-Based TOEFL Learning Transcript Dataset is a collection of textual transcripts derived from publicly available YouTube videos that provide tutorials, explanations, and learning materials related to the Test of English as a Foreign Language (TOEFL). The dataset was created to support research and educational analysis in the field of English language learning, natural language processing, and educational technology. The transcripts were obtained by converting spoken content from selected TOEFL tutorial videos into text format. The dataset includes instructional explanations, tips, practice discussions, and strategies commonly presented in TOEFL preparation videos. Each transcript represents the spoken content of a tutorial video and is organized in a structured text format to facilitate analysis. This dataset can be used for various research purposes, including language learning analysis, discourse analysis, educational content evaluation, speech-to-text research, and the development of machine learning or natural language processing models related to educational materials. All transcripts were collected from publicly accessible videos and are intended solely for research and educational purposes. The dataset does not include copyrighted video files, but only the textual transcripts generated from the spoken instructional content.
Files
Steps to reproduce
1. Video Selection - Identify publicly available YouTube videos that provide TOEFL tutorial content, tips, and practice explanations. - Ensure that videos are in English and relevant to TOEFL preparation. 2. Video Download / Access - Access the selected videos online. - (Optional) Download the videos only for offline transcription purposes, respecting YouTube’s terms of service. 3. Transcription - Convert the spoken content of each video into text using manual transcription or automated speech-to-text tools (e.g., YouTube auto-generated captions, Google Speech-to-Text, or other transcription software). 4. Data Cleaning - Remove unrelated content such as background noise, advertisements, or non-educational dialogues. - Correct grammatical errors or misrecognized words from automatic transcription tools. 5. Formatting - Organize each transcript in a structured text file format (e.g., TXT, CSV, or JSON). - Include metadata such as video title, URL, duration, and date of collection. 6. Quality Check - Review each transcript to ensure accuracy and completeness. - Compare with the original video content to verify key educational points are correctly captured. 7. Dataset Compilation - Combine all individual transcripts into a single dataset folder or file. - Provide clear file naming conventions and metadata documentation. 8. Documentation - Prepare README or dataset description including methodology, usage license, and contributor information. - Ensure the dataset can be used for research and educational purposes while respecting copyright.
Institutions
- Universitas Islam Negeri Maulana Malik IbrahimEast Java, Malang