Salient movie scenes identified by language patterns and token predictability
An Interpretable Linguistically-Grounded Analysis of Scene Saliency in Movie Screenplays
Computation and Language
Summary
This paper studies what makes certain scenes in movie screenplays stand out, or be "salient," based on how predictable and complex their language is. The authors found that instead of overall predictability of the whole scene's words, the rare and hard-to-predict words inside the scene matter most for salience. They also show salient scenes tend to have a moderate number of main characters, simpler sentence structures, more third-person narration, and abstract words about motivations. Using these insights, the authors built a model that detects important scenes better than previous methods and helps create better summaries.
What this means in practice
- •For film editors: Improve identification of key scenes in scripts for editing by using linguistic cues beyond just word counts or semantic content.
- •For natural language processing developers: Enhance screenplay summarization tools by incorporating models that combine deep language understanding with explicit linguistic features.
Authors
Uday Bindal, Ishaan Karan, Manan Garg, Ekansh Goyal, Rahul Mishra, Rajakrishnan Rajkumar
Abstract
We investigate the factors affecting salience in movie screenplay by posing two research questions: 1. {\em Does Uniform Information Density (UID) predict salient scenes?} 2. {\em What other linguistic properties distinguish salient scenes from non-salient background?} We answer both questions through a controlled corpus study on the MENSA corpus, paired with a Linguistically-Grounded Salience Model (LGSM), a dual-stream classifier that fuses a frozen BERT encoder with explicit linguistic features through adaptive gating. The UID picture is layered in long-form discourse. Per-scene UID summaries carry no independent signal once length is controlled, but at the token level salient scenes match non-salient material in the bulk while specifically suppressing the rarest, hardest-to-predict tokens in the upper tail. UID therefore lives at the within-scene token level rather than as a per-scene aggregate. Salient scenes also show a moderate and sustained cast of central characters (following an inverted-U pattern), strong third-person usage, syntactically simpler language, and more abstract vocabulary about motivation and conflict. LGSM exploits this signature and improves recall on salient scenes over a strong semantic baseline while yielding better downstream summaries.