Open intent classification is a practical yet challenging task in dialogue systems. Its objective is to accurately classify samples of known intents while at the same time detecting those of open (unknown) intents. Existing methods usually use outlier detection algorithms combined with K-class classifier to detect open intents, where K represents the class number of known intents. Different from them, in this paper, we consider another way without using outlier detection algorithms. Specifically, we directly train a (K+1)-class classifier for open intent classification, where the (K+1)-th class represents open intents. To address the challenge that training a (K+1)-class classifier with training samples of only K classes, we propose a deep model based on Soft Labeling and Manifold Mixup (SLMM). In our method, soft labeling is used to reshape the label distribution of the known intent samples, aiming at reducing model's overconfident on known intents. Manifold mixup is used to generate pseudo samples for open intents, aiming at well optimizing the decision boundary of open intents. Experiments on four benchmark datasets demonstrate that our method outperforms previous methods and achieves state-of-the-art performance. All the code and data of this work can be obtained at https://github.com/zifengcheng/SLMM.
Bootstrapping Conversational Agents With Weak Supervision
IBM Watson(IBM沃森);IBM Research AI(IBM研究院人工智能部门)
当今市场上的许多会话代理遵循标准的机器人开发框架,该框架需要训练意图分类器来识别用户输入。创建一组合适的训练样本的需求通常是开发过程中的瓶颈。在许多情况下,代理开发人员可以访问历史聊天日志,这些日志可以提供大量的训练样本。然而,用数十到数百个意图标记它们的成本通常会阻止充分利用这些聊天日志。在本文中,我们提出了一个框架称为\textit{search, label, and propagate}(SLP)的引导意图从现有的聊天日志使用弱监督。该框架通过使用搜索引擎查找示例,将数小时到数天的标记工作减少到几分钟的工作,然后依靠数据编程方法自动扩展标签。我们报告的用户研究,显示积极的用户反馈,这种新的方法来建立对话代理,并证明了使用数据编程自动标记的有效性。虽然该系统是为训练会话代理而开发的,但该框架在显着减少训练文本分类器的标记工作方面具有更广泛的应用。
IDALC: A Semi-Supervised Framework for Intent Detection and Active Learning based Correction
Department of Computer Science and Engineering, IIT Kharagpur(计算机科学与工程系,印度IIT Kharagpur);Department of Computer Science, Technische Universität Darmstadt(计算机科学系,德国Technische Universität Darmstadt);Adobe Research, Bangalore(Adobe Research,印度班加罗尔)
Multi-Task Learning of Query Intent and Named Entities using Transfer Learning
Target Corporation(塔吉特公司);AI Sciences(AI科学部门)
命名实体识别(NER)已被广泛研究,早期的算法是基于序列标记的,如隐马尔可夫模型(HMM)和条件随机场(CRF)。随后是基于神经网络的深度学习模型。最近,BERT在序列标记任务(如NER)中显示了新的最新水平的准确性。在这篇简短的文章中,我们研究了不同的方法来特定于任务的NER。任务特定的NER有两个组成部分-识别一段文本的意图(如搜索查询),然后用任务特定的命名实体标记查询。例如,我们考虑在搜索查询中标记Target商店位置的任务(可以在搜索框中输入或在Alexa或Google Home等设备中说出)。商店的位置是非常模糊的,有时很难区分,比如说一个位置和一个非位置。例如,“pickup my order at orange store”将“orange”作为商店位置,而“buy orange at target”将“orange”作为水果。我们通过多任务学习来探索这个困难,我们称之为全球到本地的信息转移。我们共同学习查询意图(即存储查找)和命名实体,通过使用多个损失函数在我们的BERT为基础的模型,并找到有趣的结果。