Across over 25,000 TV episodes, a content-aware forecasting framework finds that prior viewership dominates next-episode prediction, with NLP dialogue features adding modest and inconsistent gains across shows.
Prediction of TV program ratings based on machine learning models,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CY 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Optimizing Storytelling, Improving Audience Retention, and Reducing Waste in the Entertainment Industry
Across over 25,000 TV episodes, a content-aware forecasting framework finds that prior viewership dominates next-episode prediction, with NLP dialogue features adding modest and inconsistent gains across shows.