Indonesian Islamic Consultation Summarization Dataset

Published: 22 September 2026| Version 1 | DOI: 10.17632/7zhnrtwb4d.1
Contributors:
Sausan Shalihah Alfirdausi,

Description

This dataset contains 819 Indonesian-language Islamic consultation Q&A pairs sourced from Islamic consultation websites and transcribed Islamic lecture/Q&A videos. Each entry contains the original question-and-answer text and a reference summary. The dataset consists of five columns: id_data, which serves as a unique identifier for each data entry; title, which contains the title of the article or video; url, which provides the original source link; isi_teks, which contains the full consultation text including the question and scholarly answer and serves as the input text for summarization; and ringkasan, which contains the reference summary used as the target output for training and evaluating abstractive summarization models. The dataset can be used for fine-tuning Indonesian summarization models such as IndoT5 and mT5, benchmarking low-resource Indonesian text summarization, and conducting NLP research in the Indonesian Islamic domain.

Files

Steps to reproduce

1. Source collection: Islamic consultation texts were collected from Islamic consultation website and Islamic lecture/Q&A video transcripts from YouTube. 2. Text extraction: Article text was extracted directly from the web pages; video transcripts were obtained via automatic speech-to-text transcription of the YouTube audio using auto-generated captions, which were not yet manually cleaned. 3. Reference summary generation: For each entry, a neutral, abstractive reference summary was generated, capturing the core question and the scholarly ruling or answer. The summaries vary in length depending on the content and complexity of each source text. 4. Quality check: Checked for missing or empty text fields and spot-checked generated summaries against the source text for alignment. 5. Final compilation: Data were compiled into a single spreadsheet with columns [id_data, title, url, isi_teks, ringkasan] and exported as .xlsx.

Categories

Artificial Intelligence, Natural Language Processing

Licence