FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2285219013> ?p ?o ?g. }

• W2285219013 abstract "Identifying Metaphoric Antonyms in a Corpus Analysis of Finance Articles Aaron Gerow (gerowa@tcd.ie) School of Computer Science & Statistics, Trinity College Dublin College Green, Dublin 2, Ireland Mark T. Keane (mark.keane@ucd.ie) School of Computer Science & Informatics, University College Dublin Belfield, Dublin 4, Ireland Abstract in almost all languages. Similar spatial metaphors, of the kind we examine here, seem to ground many stock-market reports. Accounts of index, stock-market, and share move- ments tend to converge around metaphors of rising and falling, attack and retreat, gain and loss. These concepts appear to be grounded by core metaphors, with an antonymic relationship to one another, that could be glossed as GOOD IS UP and BAD IS DOWN. Lakoff and Johnson (1980) have pointed to this UP-DOWN metaphor opposition as underlying accounts of wealth (WEALTH IS UP as in high class), the rise and fall of numbers (MORE IS UP; LESS IS DOWN) and changes in quantity (CHANGE IN QUANTITY IS WAR as in retreating profits and defensive trades). In the present paper, we look at the distributive structure of these verbs’ arguments to determine whether there is empirical support for metaphoric opposites. Specifically, we try to determine whether the antonyms identified by parti- cipants in a psychological study can be shown to meaning- fully correspond to a computational analysis of the argu- ment-distributions in our corpus. Using a corpus of 17,000+ financial news reports (involving over 10M words), we perform an analysis of the argu- ment-distributions of the UP and DOWN verbs used to describe movements of indices, stocks and shares. In Study 1 participants identified antonyms of these verbs in a free-re- sponse task and a matching task from which the most commonly identified antonyms were compiled. In Study 2, we determined whether the argument-distributions for the verbs in these antonym-pairs were sufficiently similar to predict the most frequently-identified antonym. Cosine similarity correl- ates moderately with the proportions of antonym-pairs identi- fied by people (r = 0.31). More impressively, 87% of the time the most frequently-identified antonym is either the first- or second-most similar pair in the set of alternatives. The implic- ations of these results for distributional approaches to determ- ining metaphoric knowledge are discussed. Keywords: Metaphor; corpus analysis; word meaning; semantics; experimental linguistics; grounding. Introduction In recent years, significant progress has been made in deriving meaning from statistical analyses of distributions of words (Gerow & Keane, 2011a; Landauer & Dumais, 1997; Michel et al., 2010; Turney & Pantel, 2010). This distributional approach to meaning takes the view that words that occur in similar contexts tend to have similar meanings (cf. Wittgenstein, 1953) and that by analysing word usage we get at their meaning. For example, the word co-occurrence statistics derived in Latent Semantic Analysis (LSA) seem to tell us about the structure of the lexicon, as they are good predictors of reaction times in lexical decision tasks (Land- auer & Dumais, 1997). More generally, it has been suggested that significant insights into human culture and behaviour can be derived from analysing very large corpora, like the Google Books repository (Michel et al., 2010). In this paper, we apply similar distributional analyses to understand metaphoric- ally-structured knowledge underlying the antonyms between “UP and DOWN” verbs from a corpus of financial news reports. (see Gerow & Keane, 2011b, for an analysis of meta- phor hierarchies in the same data.) Lakoff (1992; Lakoff & Johnson, 1980) have argued that our understanding of many concepts, such as emotions and mental states, are grounded in a few ubiquitous metaphors. The spatial metaphors that structure emotional states – HAPPINESS IS UP and SADNESS IS DOWN – are found The Corpus In January, 2010, we carried out automated web searches that selected all articles referring to the three major world stock indices (Dow Jones, FTSE 100, and NIKKEI 225) from three websites: the New York Times (NYT, www.nyt.com), the Financial Times (FT, www.ft.com) and the British Broadcasting Corporation (BBC, www.b- bc.co.uk). These searches harvested 17,713 articles containing 10,418,266 words covering a 4-year period: January 1st, 2006 to January 1st, 2010. The by-source breakdown was FT (13,286), NYT (2,425), and BBC (2,002). The by-year breakdown was 2006 (3,869), 2007 (4,704), 2008 (5,044), 2009 (3,960), and 2010 (136). The corpus included editorials, market reports, popular pieces, and technical exposes. These three resources were chosen because they are in English and have a wide-circulation and online availability. The Financial Times made up the majority of the articles; however, the spread was actually much wider as many articles were syndicated from the Associated Press, Reuters, Bloomberg News, and Agence France-Presse. The uniqueness of the articles in the database was ensured by keying them on their first 50 characters. Once retrieved, the articles were stripped of HTML, converted to UTF-8, and shallow-parsed to extract phrasal structure using a modified version of the Apple Pie Parser" @default.
• W2285219013 created "2016-06-24" @default.
• W2285219013 creator A5016270163 @default.
• W2285219013 creator A5074878641 @default.
• W2285219013 date "2011-01-01" @default.
• W2285219013 modified "2023-09-27" @default.
• W2285219013 title "Identifying Metaphoric Antonyms in a Corpus Analysis of Finance Articles" @default.
• W2285219013 cites W1546097390 @default.
• W2285219013 cites W1662133657 @default.
• W2285219013 cites W1983578042 @default.
• W2285219013 cites W2019096529 @default.
• W2285219013 cites W2052417512 @default.
• W2285219013 cites W2098520883 @default.
• W2285219013 cites W2161325192 @default.
• W2285219013 cites W2758639250 @default.
• W2285219013 cites W58280589 @default.
• W2285219013 hasPublicationYear "2011" @default.
• W2285219013 type Work @default.
• W2285219013 sameAs 2285219013 @default.
• W2285219013 citedByCount "4" @default.
• W2285219013 countsByYear W22852190132012 @default.
• W2285219013 crossrefType "journal-article" @default.
• W2285219013 hasAuthorship W2285219013A5016270163 @default.
• W2285219013 hasAuthorship W2285219013A5074878641 @default.
• W2285219013 hasConcept C111472728 @default.
• W2285219013 hasConcept C138885662 @default.
• W2285219013 hasConcept C144024400 @default.
• W2285219013 hasConcept C15744967 @default.
• W2285219013 hasConcept C166957645 @default.
• W2285219013 hasConcept C17744445 @default.
• W2285219013 hasConcept C199539241 @default.
• W2285219013 hasConcept C2778311575 @default.
• W2285219013 hasConcept C2779343474 @default.
• W2285219013 hasConcept C2780299701 @default.
• W2285219013 hasConcept C2780668109 @default.
• W2285219013 hasConcept C41895202 @default.
• W2285219013 hasConcept C94625758 @default.
• W2285219013 hasConcept C95457728 @default.
• W2285219013 hasConceptScore W2285219013C111472728 @default.
• W2285219013 hasConceptScore W2285219013C138885662 @default.
• W2285219013 hasConceptScore W2285219013C144024400 @default.
• W2285219013 hasConceptScore W2285219013C15744967 @default.
• W2285219013 hasConceptScore W2285219013C166957645 @default.
• W2285219013 hasConceptScore W2285219013C17744445 @default.
• W2285219013 hasConceptScore W2285219013C199539241 @default.
• W2285219013 hasConceptScore W2285219013C2778311575 @default.
• W2285219013 hasConceptScore W2285219013C2779343474 @default.
• W2285219013 hasConceptScore W2285219013C2780299701 @default.
• W2285219013 hasConceptScore W2285219013C2780668109 @default.
• W2285219013 hasConceptScore W2285219013C41895202 @default.
• W2285219013 hasConceptScore W2285219013C94625758 @default.
• W2285219013 hasConceptScore W2285219013C95457728 @default.
• W2285219013 hasIssue "33" @default.
• W2285219013 hasLocation W22852190131 @default.
• W2285219013 hasOpenAccess W2285219013 @default.
• W2285219013 hasPrimaryLocation W22852190131 @default.
• W2285219013 hasRelatedWork W159327002 @default.
• W2285219013 hasRelatedWork W16501742 @default.
• W2285219013 hasRelatedWork W1670231350 @default.
• W2285219013 hasRelatedWork W1957443733 @default.
• W2285219013 hasRelatedWork W2070534650 @default.
• W2285219013 hasRelatedWork W2077652016 @default.
• W2285219013 hasRelatedWork W2129052887 @default.
• W2285219013 hasRelatedWork W2214898964 @default.
• W2285219013 hasRelatedWork W2234878719 @default.
• W2285219013 hasRelatedWork W2276861940 @default.
• W2285219013 hasRelatedWork W2282521477 @default.
• W2285219013 hasRelatedWork W2338476530 @default.
• W2285219013 hasRelatedWork W2485616202 @default.
• W2285219013 hasRelatedWork W2522306728 @default.
• W2285219013 hasRelatedWork W2603220723 @default.
• W2285219013 hasRelatedWork W284084084 @default.
• W2285219013 hasRelatedWork W2966397493 @default.
• W2285219013 hasRelatedWork W3113622281 @default.
• W2285219013 hasRelatedWork W3192057044 @default.
• W2285219013 hasRelatedWork W584048680 @default.
• W2285219013 hasVolume "33" @default.
• W2285219013 isParatext "false" @default.
• W2285219013 isRetracted "false" @default.
• W2285219013 magId "2285219013" @default.
• W2285219013 workType "article" @default.