A Systematic Classification of Knowledge, Reasoning, and Context within the ARC Dataset
This work addresses the need for better classification and annotation in the ARC dataset, which is important for researchers in AI and natural language processing, but it is incremental as it builds on prior analysis without introducing new methods.
The authors tackled the problem of unclear definitions and labeling quality for knowledge and reasoning types in the ARC dataset by proposing comprehensive definitions and analyzing the Challenge Set with ten annotators, resulting in a 42-point performance improvement for a neural machine comprehension model when using human-selected relevant sentences.
The recent work of Clark et al. introduces the AI2 Reasoning Challenge (ARC) and the associated ARC dataset that partitions open domain, complex science questions into an Easy Set and a Challenge Set. That paper includes an analysis of 100 questions with respect to the types of knowledge and reasoning required to answer them; however, it does not include clear definitions of these types, nor does it offer information about the quality of the labels. We propose a comprehensive set of definitions of knowledge and reasoning types necessary for answering the questions in the ARC dataset. Using ten annotators and a sophisticated annotation interface, we analyze the distribution of labels across the Challenge Set and statistics related to them. Additionally, we demonstrate that although naive information retrieval methods return sentences that are irrelevant to answering the query, sufficient supporting text is often present in the (ARC) corpus. Evaluating with human-selected relevant sentences improves the performance of a neural machine comprehension model by 42 points.