English

L2R2: Leveraging Ranking for Abductive Reasoning

Information Retrieval 2021-09-15 v2 Computation and Language

Abstract

The abductive natural language inference task (α\alphaNLI) is proposed to evaluate the abductive reasoning ability of a learning system. In the α\alphaNLI task, two observations are given and the most plausible hypothesis is asked to pick out from the candidates. Existing methods simply formulate it as a classification problem, thus a cross-entropy log-loss objective is used during training. However, discriminating true from false does not measure the plausibility of a hypothesis, for all the hypotheses have a chance to happen, only the probabilities are different. To fill this gap, we switch to a ranking perspective that sorts the hypotheses in order of their plausibilities. With this new perspective, a novel L2R2L2R^2 approach is proposed under the learning-to-rank framework. Firstly, training samples are reorganized into a ranking form, where two observations and their hypotheses are treated as the query and a set of candidate documents respectively. Then, an ESIM model or pre-trained language model, e.g. BERT or RoBERTa, is obtained as the scoring function. Finally, the loss functions for the ranking task can be either pair-wise or list-wise for training. The experimental results on the ART dataset reach the state-of-the-art in the public leaderboard.

Keywords

Cite

@article{arxiv.2005.11223,
  title  = {L2R2: Leveraging Ranking for Abductive Reasoning},
  author = {Yunchang Zhu and Liang Pang and Yanyan Lan and Xueqi Cheng},
  journal= {arXiv preprint arXiv:2005.11223},
  year   = {2021}
}

Comments

SIGIR 2020

R2 v1 2026-06-23T15:44:33.751Z