论文标题
切片感知神经排名
Slice-Aware Neural Ranking
论文作者
论文摘要
了解何时以及为何通过错误分析为IR任务失败的神经排名模型是研究周期的重要组成部分。在这里，我们关注（i）识别难题的类别（一对问题和回答候选者）的挑战，神经排名者无效，并且（ii）改善此类实例的神经排名。为了应对这两个挑战，我们求助于基于切片的学习，其目标是提高神经模型的效率（子集）数据的有效性。我们通过提出不同的切片函数（SF）来应对挑战（i），以选择数据集的切片 - 基于先前的工作，我们可以捕获神经排名者的不同失败。然后，对于挑战（ii），我们调整了神经排名模型以学习切片感知表示形式，即，根据模型的预测，改编的模型学会了以不同的方式表示问题和回答。我们的实验结果（源代码和数据可在https://github.com/guzpenha/slice_based_learning上获得），跨三个不同的排名任务），四个Corpora表明，基于SLICE的学习可以将其有效性提高到不含糊不清的神经排名中的平均2％。
Understanding when and why neural ranking models fail for an IR task via error analysis is an important part of the research cycle. Here we focus on the challenges of (i) identifying categories of difficult instances (a pair of question and response candidates) for which a neural ranker is ineffective and (ii) improving neural ranking for such instances. To address both challenges we resort to slice-based learning for which the goal is to improve effectiveness of neural models for slices (subsets) of data. We address challenge (i) by proposing different slicing functions (SFs) that select slices of the dataset---based on prior work we heuristically capture different failures of neural rankers. Then, for challenge (ii) we adapt a neural ranking model to learn slice-aware representations, i.e. the adapted model learns to represent the question and responses differently based on the model's prediction of which slices they belong to. Our experimental results (the source code and data are available at https://github.com/Guzpenha/slice_based_learning) across three different ranking tasks and four corpora show that slice-based learning improves the effectiveness by an average of 2% over a neural ranker that is not slice-aware.
