Narrative Modeling with Memory Chains and Semantic Supervision

TL;DR

This paper introduces a novel neural memory chain model with semantic supervision for story comprehension, achieving state-of-the-art results in story ending prediction by tracking multiple semantic aspects.

Contribution

It proposes a new memory-augmented neural network that explicitly models semantic aspects with external memory chains, enhancing story understanding.

Findings

Achieved new state-of-the-art performance on story ending prediction.

Demonstrated effectiveness of semantic supervision in neural memory models.

Outperformed competitive baselines in narrative comprehension tasks.

Abstract

Story comprehension requires a deep semantic understanding of the narrative, making it a challenging task. Inspired by previous studies on ROC Story Cloze Test, we propose a novel method, tracking various semantic aspects with external neural memory chains while encouraging each to focus on a particular semantic aspect. Evaluated on the task of story ending prediction, our model demonstrates superior performance to a collection of competitive baselines, setting a new state of the art.