Abstract
Denotation is the literal sense of a word, while connotation is its extended sense. The current paper presents a cognitive computational model of the adjective’s connotation (e.g., sweet baby). We tested the model by developing a novel algorithm – ConnoSense – that identifies the sense of an attribute’s connotation. More specifically, ConnoSense identifies the sense of an attribute such as in the case of a “sweet smile” where the attribute/adjective “sweet” is used in the sense of “friendly.” Tested on a multiple-choice test of identifying the sense of a connotation (e.g., “dark thoughts”) the algorithm gained 90% accuracy and outperforms two other models that are based on vectorial semantics. These results support the validity of our model. The paper points at the importance of fusing ideas from the semiotics interpretative tradition with experimental psychological knowledge and novel methodologies of computational semantics.
About the authors
Yair Neuman (b. 1968) is a department chair at Ben-Gurion University of the Negev. His research interests include semiotics, psychology, cultural psychology, and natural language processing. His publications include Reviving the Living: Meaning making in living systems (2008); “Automatic identification of themes in small group dynamics through the analysis of network motifs” (with D. Assaf & Y. Cohen, 2012); and “How language enables abstraction: A case study in computational cultural psychology” (with P. Turney & Y. Cohen, 2012).
Yochai Cohen (b. 1974) is the CEO of Gilasio Coding.
Dan Assaf (b. 1983) is a medical student at Tel Aviv University.
Appendix 1. The connotation test
Target adjective: WARM
Warm hug: hot (.10), wet (.11), affectionate (.32), brief (0)
warm relations: loving (.47), tense (.15), public (.14), amicable (.20)
warm welcome: affectionate (.31), adoring (.20), Royal (.09), tense (.15)
warm feelings: comforting (.25), funny (.20), pleasant (.21), mixed (.12)
warm personality: friendly (.41), split (.09), strong (.22), antisocial (.13)
warm greeting: friendly (.43), formal (.14), traditional (.16), usual (.17)
warm support: ardent (.23), public (.15), social (.14), financial (.13)
warm debate: lively (.50), public (.14), current (.13), big (.13)
warm smile: friendly (.42), big (.14), sly (.17), quick (.13)
warm memories: lovable (.22), vivid (.21), old (.12), long (.13)
warm applause: affectionate (.32), polite (.26), quiet (.21), mild (.13)
Target adjective: SWEET
sweet dream: good (.20), golden (.12), wonderful (.25), vivid (.20)
sweet revenge: gratifying (.21), political (.14), passionate (.24), ultimate (.16)
sweet smile: gentle (.49), big (.15), sugary (.15), strange (.19)
sweet baby: cute (.44), little (0), beautiful (.24), honeyed (.14)
sweet music: pleasing (.24), classical (.14), loud (.16), choral (.13)
sweet song: pleasing (.24), old (.14), popular (.15), sad (.19)
sweet child: lovable (.23), young (.17), sick (.16), small (.15)
sweet deal: satisfying (.21), fair (.17), bigger (.16), new (.12)
sweet voice: melodic (.20), male (.12), deep (.17), inner (0)
sweet memories: pleasant (.44), vivid (.20), old (.13), long (.14)
sweet charity: gratifying (.21), Christian (.13), local (.12), Muslim (.13)
sweet victory: pleasant (.40), upset (.17), great (.19), electoral (.08)
Target adjective: DEEP
deep depression: intense (.50), terrible (.21), recurrent (.13), great (.18)
deep sleep: profound (.50), feigned (.19), archetypal (.18), cold (.14)
deep trouble: serious (.21), enormous (.20), powerful (.21), financial (.13)
deep understanding: profound (.51), tacit (.19), (broad (.18), collective (.18)
deep sigh: profound (.49), collective (.17), small (.13), audible (.16)
deep emotion: strong (.21), bottomless (.14), human (.15), real (.16)
deep thoughts: profound (.52), private (.13), random (.08), happy (.14)
deep happiness: intense (.49), future (.16), human (.16), martial (.09)
deep voice: low (.10), female (.11), human (.14), male (.10)
deep secret: mysterious (.49), guarded (.15), soviet (.13), open (.11)
deep theory: abstruse (.19), low (.10), economic (.15), cognitive (.16)
deep lecture: recondite (.18), public (.13), large (.14), free (.10)
Target adjective: DARK
dark comedy: gallous (.12), broad (.10), classic (.10), black (.13)
dark humor: wicked (.20), light (.13), Jewish (.13), ethnic (.10)
dark secret: concealed (.17), little (0), soviet (.12), open (.09)
dark thoughts: evil (.57), intimate (.15), romantic (.12), intrusive (.14)
dark purpose: evil (.54), sole (.09), general (.11), main (.07)
dark vision: pessimistic (.144), clear (.1409), personal (.13), comic (.09)
dark look: threatening (.50), close (.10), fresh (.07), curios (.16)
dark deeds: sinister (.22), good (.14), heroic (.18), past (.14)
dark angel: vicious (.19), stone (.11), guardian (.09), Christmas (.10)
dark mood: gloomy (.20), good (.14), positive (.13), public (.09)
dark forces: evil (.57), armed (.12), social (.09), economic (.10)
dark magic: sinister (.23), practical (.13), ancient (.15), powerful (.17)
Target adjective: HARD
hard task: difficult (.49), solid (.11), simple (.18), joint (.13)
hard look: dispassionate (.16), brief (0), close (.17), firm (.10)
hark bargainer: dispassionate (.16), solid (.10), good (.15), trustworthy (.16)
hard winds: strong (.22), high (.12), political (.16), cold (.11)
hard blow: strong (.22), firm (.11), low (.11), single (.13)
hard liquors: strong (.22), flavored (0), organic (.08), exotic (.12)
hard drinker: intemperate (.16), young (.13), social (.14), oldest (0)
hard luck: tough (.18), good (.16), pure (.11), random 9.07)
hard feelings: negative (.17), raw (.09), bad (.13), awful (.16)
hard evidence: convincing (.20), scant (.14), recent (.17), criminal (.17)
hard time: difficult (.49), insufficient (.16), wasted (.13), tough (.18)
hard work: difficult (.48), tough (.18), extra (.09), diligent (.17)
References
Assaf, Dan, YairNeuman & YochaiCohen, 2013. A cognitively motivated word sense induction algorithm. Paper presented at the IEEE SSCI Computational Intelligence Conference, Singapore.Search in Google Scholar
Allan, Keith.2007. The pragmatics of connotation. Journal of Pragmatics39. 1047–1057.10.1016/j.pragma.2006.08.004Search in Google Scholar
Andrews, Mark, GabriellaVigliocco & DavidVinson.2009. Integrating experiential and distributional data to learn semantic representations. Psychological Review116. 463–498.10.1037/a0016261Search in Google Scholar
Arunachalam, Sudha & Sandra R.Waxman.2010. Language and conceptual development. Cognitive Science1(4). 548–558.10.1002/wcs.37Search in Google Scholar
Binder, Jeffrey R. & Rutvik H.Desai. 2011. The neurobiology of semantic memory. Trends in Cognitive Sciences15. 527–536.10.1016/j.tics.2011.10.001Search in Google Scholar
Bringsjord, Selmer & BettinaSchimanski.2003. What is artificial intelligence? Psychometric AI as an answer. International Joint Conference on Artificial Intelligence18. 887–893.Search in Google Scholar
Brownell, Hiram H., Heather H.Potter, DianeMichelow & HowardGardner.1984. Sensitivity to lexical denotation and connotation in brain-damaged patients: A double dissociation?Brain and Language22. 253–265.10.1016/0093-934X(84)90093-2Search in Google Scholar
Corrigan, Roberta.2004. The acquisition of word connotations: Asking “What happened?”Journal of Child Language31. 381–398.10.1017/S0305000903005981Search in Google Scholar
Danesi, Marcel.2003. Metaphorical “networks” and verbal communication. Sign Systems Studies31. 341–363.10.12697/SSS.2003.31.2.02Search in Google Scholar
Davies, Mark.2009. The 385+ million word corpus of contemporary American English (19902008+): Design, architecture, and linguistic insights. International Journal of Corpus Linguistics14. 159–190.Search in Google Scholar
Just, Marcel Adam, Vladimir L.Cherkassky, SandeshAryal & Tom M.Mitchell. 2010. A neurosemantic theory of concrete noun representation based on the underlying brain codes. PloS One5(1). e8622.10.1371/journal.pone.0008622Search in Google Scholar
Krishnakumaran, Saisuresh & XiaojinZhu.2007. Hunting elusive metaphors using lexical resources. In Proceedings of the workshop on computational approaches to figurative language, 13–20. Stroudsburg: Association for Computational Linguistics.10.3115/1611528.1611531Search in Google Scholar
Lakoff, George & MarkJohnson.1999. Philosophy in the flesh. New York: Basic.Search in Google Scholar
Light, Marc & WarrenGreiff.2002. Statistical models for the induction and use of selectional preferences. Cognitive Science26. 269–281.10.1207/s15516709cog2603_4Search in Google Scholar
Neuman, Yair, DanAssaf, YohaiCohen, MarkLast, ShlomoArgamon, NewtonHoward & OphirFrieder.2013a. Metaphor identification in large texts corpora. PloS One8(4). e62343.10.1371/journal.pone.0062343Search in Google Scholar
Neuman, Yair, DanAssaf & YohaiCohen.2013b. Fusing distributional and experiential information for measuring semantic relatedness. Information Fusion14(3). 281–287.10.1016/j.inffus.2012.02.001Search in Google Scholar
Rosch, Eleanor H.1973. On the internal structure of perceptual and semantic categories. In T.Moore (ed.), Cognitive development and acquisition of language, 111–144. New York: Academic Press.Search in Google Scholar
Rosch, Eleanor H.2011. “Slow lettuce”: Categories, concepts, fuzzy sets, and logical deduction. In R.Belohlavek & G. L.Klir (eds.), Concepts and fuzzy logic, 89–121. Cambridge, MA: MIT Press.Search in Google Scholar
Sebeok, Thomas A. & MarcelDanesi. 2000. The forms of meaning: Modeling systems theory and semiotics. Berlin: Mouton de Gruyter.10.1515/9783110816143Search in Google Scholar
Turney, Peter D.2012. Domain and function: A dual-space model of semantic relations and compositions. Journal of Artificial Intelligence Research44. 533–585.10.1613/jair.3640Search in Google Scholar
Turney, Peter D. & PatrickPantel. 2010. From frequency to meaning: Vector space models of semantics. Journal of Artificial Intelligence Research37. 141–188.10.1613/jair.2934Search in Google Scholar
Turney, Peter D., YairNeuman, YochaiCohen & DanAssaf.2011. Literal and metaphorical sense identification through concrete and abstract context. In Proceedings of the 2011 conference on the empirical methods in natural language processing, 680–690. Edinburgh.Search in Google Scholar
Waxman, Sandra R. & Raquel S.Klibanoff. 2000. The role of comparison in the extension of novel adjectives. Developmental Psychology36. 571–581.10.1037/0012-1649.36.5.571Search in Google Scholar
Waxman, Sandra R. & Jeffrey L.Lidz. 2006. Early world learning. In D.Kuhn & R.Siegler (eds.), Handbook of child psychology, vol. 2, 299–335. Hoboken, NJ: Wiley.10.1002/9780470147658.chpsy0207Search in Google Scholar
©2015 by De Gruyter Mouton
