Better Document-level Sentiment Analysis from RST Discourse Parsing

Abstract

Discourse parsing is an integral part of understanding information flow and argumentative structure in documents. Most previous research has focused on inducing and evaluating models from the English RST Discourse Treebank. However, discourse treebanks for other languages exist , including Spanish, German, Basque, Dutch and Brazilian Portuguese. The tree-banks share the same underlying linguistic theory, but differ slightly in the way documents are annotated. In this paper, we present (a) a new discourse parser which is simpler, yet competitive (significantly better on 2/3 metrics) to state of the art for English, (b) a harmonization of discourse treebanks across languages, enabling us to present (c) what to the best of our knowledge are the first experiments on cross-lingual discourse parsing.

Cross-lingual RST Discourse Parsing

“…This theory guided the annotation of the RST Discourse Treebank (RST-DT) for English, from which several textlevel discourse parsers have been induced (Hernault et al, 2010;Joty et al, 2012;Feng and Hirst, 2014;Li et al, 2014;Ji and Eisenstein, 2014). Such parsers have proven to be useful for various downstream applications (Daumé III and Marcu, 2009;Burstein et al, 2003;Higgins et al, 2004;Thione et al, 2004;Sporleder and Lapata, 2005;Taboada and Mann, 2006;Louis et al, 2010;Bhatia et al, 2015).…”

Exploring Joint Neural Model for Sentence Level Discourse Parsing and Sentiment Analysis

“…In the Positive/Negative setting, if the prediction and the target had the same sign, they were considered equal. Notice that this is different from training a classifier for binary classification, which is a much easier task (see (Bhatia et al, 2015)). The difference in accuracy between these two settings signals that distinguishing between very positive and positive and distinguishing between very negative and negative is rather hard.…”

Improving Sentiment Analysis with Document-Level Semantic Relationships from Rhetoric Discourse Structures

“…However, these approaches usually rely exclusively on the number of occurrences of certain relevant words or phrases. As such, these methods are not capable of taking into account the actual semantic relationships between parts of the document, individual sentences or even subclauses [2], [3]. Accordingly, these methods often struggle to achieve a favorable performance for longer documents and, hence, new approaches are desired [4].…”

“…Proceedings of the 50th Hawaii International Conference on System Sciences | 2017 URI: http://hdl.handle.net/10125/41288 ISBN: 978-0-9981331-0-2 CC-BY-NC-ND The only investigation of machine learning for rhetoricstructure-based sentiment analysis is given in [2], which utilizes recursive neural networks (RNN). As such, it struggles with small datasets, as in this case, where we need to rely on hand-crafted feature engineering.…”