Structured Sentiment Analysis as Dependency Graph Parsing

Abstract

Structured sentiment analysis attempts to extract full opinion tuples from a text, but over time this task has been subdivided into smaller and smaller sub-tasks, e.g., target extraction or targeted polarity classification. We argue that this division has become counterproductive and propose a new unified framework to remedy the situation. We cast the structured sentiment problem as dependency graph parsing, where the nodes are spans of sentiment holders, targets and expressions, and the arcs are the relations between them. We perform experiments on five datasets in four languages (English, Norwegian, Basque, and Catalan) and show that this approach leads to strong improvements over state-of-the-art baselines. Our analysis shows that refining the sentiment graphs with syntactic dependency information further improves results.

Direct parsing to sentiment graphs

“…Proposing a dependency parsing approach to the full task of SSA, Barnes et al (2021) show that it leads to strong improvements over state-of-the-art baselines. Peng et al (2021) propose a sparse fuzzy attention mechanism to deal with the sparseness of dependency arcs in the models from Barnes et al (2021) and show further improvements.…”

“…Dependency parsing approaches have recently shown promising results for SSA (Barnes et al, 2021;Peng et al, 2021). Here we present a novel sentiment parser which, unlike previous attempts, predicts sentiment graphs directly from text without reliance on heuristic lossy conversions to intermediate dependency representations.…”

Sparse Fuzzy Attention for Structured Sentiment Analysis

“…For comparison with previous state-of-the-art model (Barnes et al, 2021), we conduct our experiments on datasets of multiple languages, including hotel reviews MultiB EU , MultiB CA (Barnes et al, 2018) in Basque and Catalan, professional reviews NoReC Fine (Øvrelid et al, 2020) in Norwegian, news wire text MPQA (Wiebe et al, 2005) in English and reviews of online universities and ecommerce DS Unis (Toprak et al, 2010) in English.…”

“…Thus, (Barnes et al, 2021) propose a parsingbased procedure which implements a real unified structured sentiment analysis, where spans, relations and labels are all cast into dependencies between components in context. Our attention mechanism refines this procedure by adapting the parser for rather denser real dependency graphs to sparser and more continuous formulated dependency graphs from structured sentiment analysis.…”

“…A unified scenario is recently proposed for structured sentiment analysis and leads to much improvement in model performance. Barnes et al (2021) model structured sentiment analysis as dependency parsing and applies a parser from dependency parsing for the formulated task as in Figure 1. More related works can be referred to Appendix A.…”