Rhetorical Structure Theory for polarity estimation: An experimental study
Sentiment analysis tools often rely on counts of sentiment-carrying words, ignoring structural aspects of content. Natural Language Processing has been fruitfully exploited in text mining, but advanced discourse processing is still nonpervasive for mining opinions. Some studies, however, extracted opinions based on the discursive role of text segments. The merits of such computationally intensive analyses have thus far been assessed in very specific, small-scale scenarios. In this paper, we investigate the usefulness of Rhetorical Structure Theory in various sentiment analysis tasks on different types of information sources. First, we demonstrate how to perform a large-scale ranking of individual blog posts in terms of their overall polarity, by exploiting the rhetorical structure of a few key evaluative sentences. In order to further validate our findings, we additionally explore the potential of Rhetorical Structure Theory in sentence-level polarity classification of news and product reviews. Our most valuable polarity classification features turn out to capture the way in which polar terms are used, rather than the sentiment-carrying words per se.
|Keywords||Information, Natural Language Processing, Retrieval, Rhetorical Structure Theory, Sentiment analysis, Text mining|
|Persistent URL||dx.doi.org/10.1016/j.datak.2014.07.009, hdl.handle.net/1765/101539|
|Series||ERIM Top-Core Articles|
|Journal||Data and Knowledge Engineering|
Chenlo, J.M, Hogenboom, A.C, & Losada, D.E. (2014). Rhetorical Structure Theory for polarity estimation: An experimental study. Data and Knowledge Engineering (Vol. 94, pp. 135–147). doi:10.1016/j.datak.2014.07.009