Weighted Global Normalization for Multiple Choice Reading Comprehension over Long Documents
This addresses the fragility of span prediction models for answer selection in long-document reading comprehension, though it appears incremental as it builds on an adapted span prediction model.
The paper tackles the problem of multiple choice reading comprehension over long documents by introducing weighted global normalization of predictions, achieving a +36.2 improvement in Mean Reciprocal Rank on the NarrativeQA dataset.
Motivated by recent evidence pointing out the fragility of high-performing span prediction models, we direct our attention to multiple choice reading comprehension. In particular, this work introduces a novel method for improving answer selection on long documents through weighted global normalization of predictions over portions of the documents. We show that applying our method to a span prediction model adapted for answer selection helps model performance on long summaries from NarrativeQA, a challenging reading comprehension dataset with an answer selection task, and we strongly improve on the task baseline performance by +36.2 Mean Reciprocal Rank.