CogME: A Cognition-Inspired Multi-Dimensional Evaluation Metric for Story Understanding
This work addresses the need for more sophisticated evaluation metrics in AI story understanding, offering a framework that aligns with human cognitive processes to guide development in higher cognitive functions.
The authors tackled the problem of evaluating AI models for story understanding by introducing CogME, a cognition-inspired multi-dimensional metric, which provides nuanced insights into model strengths and weaknesses and dataset characteristics, as demonstrated in a case study with the DramaQA dataset.
We introduce CogME, a cognition-inspired, multi-dimensional evaluation metric designed for AI models focusing on story understanding. CogME is a framework grounded in human thinking strategies and story elements that involve story understanding. With a specific breakdown of the questions, this approach provides a nuanced assessment revealing not only AI models' particular strengths and weaknesses but also the characteristics of the benchmark dataset. Our case study with the DramaQA dataset demonstrates a refined analysis of the model and the benchmark dataset. We argue the need for metrics based on understanding the nature of tasks and designed to align closely with human cognitive processes. This approach provides insights beyond traditional overall scores and paves the way for more sophisticated AI development targeting higher cognitive functions.