Better Document-level Sentiment Analysis from RST Discourse Parsing
This work addresses sentiment analysis for documents by using discourse structure, offering incremental improvements in accuracy.
The paper tackled document-level sentiment analysis by leveraging Rhetorical Structure Theory (RST) discourse parsing, showing that reweighting discourse units improved lexicon-based methods and a recursive neural network over RST structure outperformed classification-based approaches with significant gains.
Discourse structure is the hidden link between surface features and document-level properties, such as sentiment polarity. We show that the discourse analyses produced by Rhetorical Structure Theory (RST) parsers can improve document-level sentiment analysis, via composition of local information up the discourse tree. First, we show that reweighting discourse units according to their position in a dependency representation of the rhetorical structure can yield substantial improvements on lexicon-based sentiment analysis. Next, we present a recursive neural network over the RST structure, which offers significant improvements over classification-based methods.