Topic detection using paragraph vectors to support active learning in systematic reviews.
Topic detection using paragraph vectors to support active learning in systematic reviews.
复制标题
DOI：
10.1016/j.jbi.2016.06.001
复制
发表时间：
2016-08
影响因子：
4.5
通讯作者：
Ananiadou S
中科院分区：
文献类型：
--
作者：
Hashimoto K;Kontonatsios G;Miwa M;Ananiadou S
提出了一种基于段落向量的话题检测方法。该方法与主动学习者相结合，加快了引文筛选。当应用于临床和公共卫生审查时，该方法的性能优于LDA。系统评审需要专家评审员手动筛选数千条引文，以确定与评审有关的所有文章。主动学习文本分类是一种有监督的机器学习方法，通过将系统评论的引文筛选过程半自动进行，已被证明可以显著减少人工标注工作量。在本文中，我们提出了一种新的主题检测方法，该方法引入了学习的信息表示，以提高潜在的主动学习者的表现。我们提出的主题检测方法使用基于神经网络的向量空间模型来捕捉文档之间的语义相似性。我们首先在向量空间中表示文档，然后将文档聚类成预定数量的聚类。将聚类的质心视为潜在主题。然后，我们将每个文档表示为潜在主题的混合。出于评估目的，我们使用了主动学习策略，使用了我们新的主题检测方法和基线主题模型(即潜在Dirichlet分配)。结果表明，与基线方法相比，我们的方法能够实现对符合条件的研究的高灵敏度，并且显著降低了人工标注成本。这一观察结果在两次临床审查和三次公共卫生审查中是一致的。本文中介绍的工具可从https://nactem.ac.uk/pvtopic/.获得
We propose a topic detection method based on paragraph vectors. The method is integrated with an active learner to accelerate citation screening. The method outperforms LDA when applied to clinical and public health reviews. Systematic reviews require expert reviewers to manually screen thousands of citations in order to identify all relevant articles to the review. Active learning text classification is a supervised machine learning approach that has been shown to significantly reduce the manual annotation workload by semi-automating the citation screening process of systematic reviews. In this paper, we present a new topic detection method that induces an informative representation of studies, to improve the performance of the underlying active learner. Our proposed topic detection method uses a neural network-based vector space model to capture semantic similarities between documents. We firstly represent documents within the vector space, and cluster the documents into a predefined number of clusters. The centroids of the clusters are treated as latent topics. We then represent each document as a mixture of latent topics. For evaluation purposes, we employ the active learning strategy using both our novel topic detection method and a baseline topic model (i.e., Latent Dirichlet Allocation). Results obtained demonstrate that our method is able to achieve a high sensitivity of eligible studies and a significantly reduced manual annotation cost when compared to the baseline method. This observation is consistent across two clinical and three public health reviews. The tool introduced in this work is available from https://nactem.ac.uk/pvtopic/.