This repo presents data illustrating results obtained by applying Multi Model Mass Media Assessment (M4A) to a corpora of news publication from Kazakhstan media:
Data:
- "Top news" folder contains lists of top news by corresponding criteria
- "Validation" folder contains files with random datasets of 1000 news. final_class.xlsx, resonance.xlsx, sentiment.xlsx and social.xlsx contains news and their assesment by the propsed model, while corresponding files with _blind_labelled postfix contain experts labelling accoring to Likert scale. Resonance_truth contains popularity estimation based on actual number of views.
- Corpora of news publication from 40 Kazakhstani sources published from 01.01.2018 to 31.12.2019 can be downloaded here