OCanada-TVNews: A Two-Year Dataset for Gender and Topic Diversity in Canadian TV News Broadcasts

Abstract

This paper presents OCanada-TVNews, a two-year dataset of Canadian TV news broadcasts, from four major channels: CBC The National, CTV The National, Global News, and RadioCanada Info, retrieved from channels’ YouTube playlists. The dataset consists of 3982 videos, in total 999 hours. Each video is processed through a systematic pipeline that involves audio extraction, speaker diarization, gender detection, and topic classification. In addition, for each video, metadata containing user engagements is also provided. By merging speaker segments with gender labels and topic categories, this dataset offers a detailed view of who speaks and what is discussed on a daily basis. An exploratory analysis highlights persistent disparities in gender representation and provides insights into the topics that dominate coverage. OCanada-TVNews serves as a basis for a further study of media diversity, content trends, and potential biases in broadcast journalism.