Enhancing a Lexicon of Polarity Shifters through the Supervised Classification of Shifting Directions

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pages 5010–5016 Marseille, 11–16 May 2020 c⃝ European Language Resources Association (ELRA), licensed under CC-BY-NC 5010 Enhancing a Lexicon of Polarity Shifters through the Supervised Classification of Shifting Directions Marc Schulder*†, Michael Wiegand*‡ and Josef Ruppenhofer◦ * Spoken Language Systems, Saarland Informatics Campus, Saarland University, Germany † Institute for German Sign Language, University of Hamburg, Germany ‡ Leibniz ScienceCampus, Heidelberg/Mannheim, Germany ◦ Institute for German Language, Mannheim, Germany marc.schulder@uni-hamburg.de, wiegand@ids-mannheim.de, ruppenhofer@ids-mannheim.de Abstract The sentiment polarity of an expression (whether it is perceived as positive, negative or neutral) can be influenced by a number of phenomena, foremost among them negation. Apart from closed-class negation words like no, not or without, negation can also be caused by so-called polarity shifters. These are content words, such as verbs, nouns or adjectives, that shift polarities in their opposite direction, e. g. abandoned in “abandoned hope” or alleviate in “alleviate pain”. Many polarity shifters can affect both positive and negative polar expressions, shifting them towards the opposing polarity. However, other shifters are restricted to a single shifting direction. Recoup shifts negative to positive in “recoup your losses”, but does not affect the positive polarity of fortune in “recoup a fortune”. Existing polarity shifter lexica only specify whether a word can, in general, cause shifting, but they do not specify when this is limited to one shifting direction. To address this issue we introduce a supervised classifier that determines the shifting direction of shifters. This classifier uses both resource-driven features, such as WordNet relations, and data-driven features like in-context polarity conflicts. Using this classifier we enhance the largest available polarity shifter lexicon. Keywords: Sentiment Analysis; Polarity Shifter; Negation; Lexical Semantics; Lexicon; Supervised Classification 1. Introduction Polarity shifters are content words that have semantic prop- erties similar to those of negation words like not, no or without. Like negation words, they can change the sen- timent polarity of an expression (also known as valence) by either inverting it or reducing its intensity. For exam- ple, the negated statement in (1) can also be expressed by the verbal polarity shifter fail or its nominal and adjectival equivalents, as seen in (2)–(4). (1) Peter [did notnegation [pass the exam]+]−. (2) Peter [failedshifter to [pass the exam]+]−. (3) Peter’s [failureshifter to [pass the exam]+]− (4) Peter’s [failedshifter [attempt to pass the exam]+]− Many polarity shifters can affect both positive and negative expressions. In (5), the verbal shifter destroy shifts a posi- tive polar expression to negative, while in (6) it shifts from negative to positive. (5) It [destroyedshifter their [hopes]+]−. (6) The medication will [destroyshifter the [cancer]−]+. Other shifters, however, are unidirectional and only shift polarities in one direction (Wilson et al., 2005). The verbal shifter to risk, for example, shifts only positive polar ex- pressions like good health in (7), while the polarity of neg- ative polar expressions like war in (8) remains unaffected. Similarly, the adjectival shifter antiquated shifts the posi- tive noun ideal in (9), …

Content truncated to 500 words