CLICK-ID: A Novel Dataset for Indonesian Clickbait Headlines

Published: 11 May 2020| Version 1 | DOI: 10.17632/k42j7x2kpn.1
Contributors:
Andika William,

Description

The CLICK-ID dataset is a collection of Indonesian news headlines that was collected from 12 local online news publishers; detikNews, Fimela, Kapanlagi, Kompas, Liputan6, Okezone, Posmetro-Medan, Republika, Sindonews, Tempo, Tribunnews, and Wowkeren. This dataset is comprised of mainly two parts; (i) 46,119 raw article data, and (ii) 15,000 clickbait annotated sample headlines. Annotation was conducted with 3 annotator examining each headline. Judgment were based only on the headline. The majority then is considered as the ground truth. In the annotated sample, our annotation shows 6,290 clickbait and 8,710 non-clickbait. There are 2 folder; raw and annotated. The raw folder contains the original files that were collected from the publisher's websites. On the other hand, the annotated folder contains the 15,000 chosen sample of annotated headlines. Both folders contains three folders with names corresponding to their contents' file extensions; (.csv) and (.xlsx), plus one combined folder. The raw data files contains the attributes: title, source, date, time, category, sub-category, content, and url. While the annotated data files contains the attributes: title, label, and label_score.

Files

Institutions

Universitas Gadjah Mada

Categories

Natural Language Processing, Machine Learning, Indonesian Language, Categorization, Digital Media

Licence