arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2201.03356cs.IRcs.LG
收藏

神经信息检索中长主题序列的持续学习

Continual Learning of Long Topic Sequences in Neural Information Retrieval

CNRS-ISIR, Sorbonne University(法国国家科学研究中心-智能系统与机器人研究所,索邦大学)

Thomas Gerald, Laure Soulier

摘要:

在信息检索(IR)系统中,趋势和用户的兴趣可能随时间改变,从而改变请求的分布或要推荐的内容。由于神经排序方法严重依赖于训练数据,因此了解最近IR方法的传输能力以解决长期的新领域是至关重要的。在本文中,我们首先提出了一个数据集的基础上MSMarco语料库,旨在建模一个很长的主题流以及IR属性驱动的控制设置。然后,我们深入分析最近的神经IR模型的能力,同时不断学习这些流。我们的实证研究强调了在哪些特定情况下会发生灾难性遗忘(例如,任务之间的相似性水平、文本长度的特殊性以及学习模型的方式),以提供模型设计方面的未来方向。

英文摘要:

In information retrieval (IR) systems, trends and users' interests may change over time, altering either the distribution of requests or contents to be recommended. Since neural ranking approaches heavily depend on the training data, it is crucial to understand the transfer capacity of recent IR approaches to address new domains in the long term. In this paper, we first propose a dataset based upon the MSMarco corpus aiming at modeling a long stream of topics as well as IR property-driven controlled settings. We then in-depth analyze the ability of recent neural IR models while continually learning those streams. Our empirical study highlights in which particular cases catastrophic forgetting occurs (e.g., level of similarity between tasks, peculiarities on text length, and ways of learning models) to provide future directions in terms of model design.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑