A Span-based Linearization for Constituent Trees

Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 313–322 July 5 - 10, 2020. c⃝2020 Association for Computational Linguistics 313 A Joint Model for Document Segmentation and Segment Labeling Joe Barrow∗ University of Maryland jdbarrow@cs.umd.edu Rajiv Jain Adobe Research rajijain@adobe.com Vlad I. Morariu Adobe Research morariu@adobe.com Varun Manjunatha Adobe Research vmanjuna@adobe.com Douglas W. Oard University of Maryland oard@umd.edu Philip Resnik University of Maryland resnik@umd.edu Abstract Text segmentation aims to uncover latent struc- ture by dividing text from a document into co- herent sections. Where previous work on text segmentation considers the tasks of document segmentation and segment labeling separately, we show that the tasks contain complemen- tary information and are best addressed jointly. We introduce the Segment Pooling LSTM (S- LSTM) model, which is capable of jointly seg- menting a document and labeling segments. In support of joint training, we develop a method for teaching the model to recover from errors by aligning the predicted and ground truth seg- ments. We show that S-LSTM reduces seg- mentation error by 30% on average, while also improving segment labeling. 1 Introduction A well-written document is rich not only in con- tent but also in structure. One type of structure is the grouping of content into topically coherent segments. These segmented documents have many uses across various domains and downstream tasks. Segmentation can, for example, be used to con- vert unstructured medical dictations into clinical reports (Sadoughi et al., 2018), which in turn can help with medical coding (since a diagnosis men- tioned in a "Medical History" might be different from a diagnosis mentioned in an "Intake" sec- tion (Ganesan and Subotin, 2014)). Segmentation can also be used downstream for retrieval (Hearst and Plaunt, 2002; Edinger et al., 2017; Allan et al., 1998), where it can be particularly useful when applied to informal text or speech that lacks ex- plicit segment markup. Topically segmented docu- ments are also useful for pre-reading (the process of skimming or surveying a text prior to careful reading), thus serving as an aid for reading compre- hension (Swaffar et al., 1991; Ajideh, 2003). ∗⋆ Work done while interning at Adobe. Uncovering latent, topically coherent segments of text is a difficult problem because it requires solving a chicken-and-egg problem: determining the segment topics is easier if segment boundaries are given, and identifying the boundaries of seg- ments is easier if the topic(s) addressed in parts of the document are known. Prior approaches to text segmentation can largely be split into two cate- gories that break the cycle by sequentially solving the two problems: those that attempt to directly predict segment bounds (Koshorek et al., 2018), and those that attempt to predict topics per passage (e.g., per sentence) and use measures of coherence for post hoc segmentation (Hearst, 1997; Arnold et al.; Eisenstein and Barzilay, 2008; Riedl and Biemann, 2012; Glavaš et al., 2016). The benefit of the topic modeling approach is that it can work in unsupervised settings where collecting ground truth segmentations is …

Content truncated to 500 words