Introducing a Lexicon of Verbal Polarity Shifters for English

Introducing a Lexicon of Verbal Polarity Shifters for English Marc Schulder*, Michael Wiegand*, Josef Ruppenhofer†, Stephanie K¨oser* * Spoken Language Systems, Saarland University, Germany † Institute for German Language, Mannheim, Germany marc.schulder@lsv.uni-saarland.de, michael.wiegand@lsv.uni-saarland.de, ruppenhofer@ids-mannheim.de, skoeser@coli.uni-saarland.de Abstract The sentiment polarity of a phrase does not only depend on the polarities of its words, but also on how these are affected by their context. Negation words (e.g. not, no, never) can change the polarity of a phrase. Similarly, verbs and other content words can also act as polarity shifters (e.g. fail, deny, alleviate). While individually more sparse, they are far more numerous. Among verbs alone, there are more than 1200 shifters. However, sentiment analysis systems barely consider polarity shifters other than negation words. A major reason for this is the scarcity of lexicons and corpora that provide information on them. We introduce a lexicon of verbal polarity shifters that covers the entirety of verbs found in WordNet. We provide a fine-grained annotation of individual word senses, as well as information for each verbal shifter on the syntactic scopes that it can affect. Keywords: Sentiment Analysis, Polarity Shifter, Negation, Lexical Semantics, Lexicon 1. Introduction Polarity shifters are content words that exhibit semantic properties similar to negation. For example, the negated statement in (1) can also be achieved by the verbal shifter fail instead of the negation not, as shown in (2). (1) Peter did not pass the exam. (2) Peter failedshifter to pass the exam. As with negation words, polarity shifters change the polar- ity of a statement. This can happen to both positive and negative statements. In (3) the positive polarity of scholar- ship is shifted by denied, resulting in a negative polarity for the phrase. Conversely, the overall polarity of (4) is positive despite the negative polarity of pain. (3) She was [deniedshifter the [scholarship]+]−. (4) The new treatment has [alleviatedshifter her [pain]−]+. Polarity shifting is also caused by other content word classes, such as nouns (e.g. downfall) and adjectives (e.g. devoid). However, this work focusses on verbs, due to their importance as minimal semantic units, far-reaching scopes and potential basis for nominal shifter lexicons (see §2.2.). Knowledge of polarity shifting is important for a variety of tasks, especially sentiment analysis (Wiegand et al., 2010; Liu, 2012; Wilson et al., 2005), as well as relation extrac- tion (Sanchez-Graillet and Poesio, 2007) and textual entail- ment recognition (Harabagiu et al., 2006). The majority of research into polarity shifting for sentiment analysis has focussed on negation words (Wiegand et al., 2010; Schouten and Frasincar, 2016; Pak and Paroubek, 2010). Negation words (e.g. not, no, never) are mostly function words, of which only a small number exists, so ex- haustive coverage is comparatively simple. Content word classes, such as verbs, are considerably more difficult to cover comprehensively due to their sheer number. For ex- ample, WordNet (Miller et al., 1990) contains over 10k verbal lemmas. Most verbs are also far less frequent than common negation words, making individual verbal shifters seem less important. However, …

Content truncated to 500 words