SCIENTIA SINICA Informationis, Volume 49 , Issue 8 : 1005-1018(2019) https://doi.org/10.1360/N112018-00266

Knowledge representation of non-literal meanings of Chinese words and its applications

More info
  • ReceivedOct 8, 2018
  • AcceptedDec 26, 2018
  • PublishedAug 9, 2019


Funded by



[1] Yuan Y L. Linguistic assumptions on natural language understanding. Soc Sci China, 1993, 1: 189--206. Google Scholar

[2] Chao Y. A Grammar of Spoken Chinese. Oakland: University of California Press, 1968. Google Scholar

[3] Liu Y, Lin Z, Kang S C. Towards a description of Chinese morphemic concepts and semantic word-formation. J Chin Inf Proc, 2018, 32: 12--21. Google Scholar

[4] Yu S W, Wang Z M, Zhu X F. Literary language and natural language understanding. In: Proceedings of the 25th Anniversary Academic Conference of Chinese Information Society of China. Beijing: Tsinghua University Press, 2006. 72--79. Google Scholar

[5] Bloomfield L. Language. New York: Henry Holt, 1933. Google Scholar

[6] Ullmann S. The Principles of Semantics. Glasgow: Jackson, Son & Co., 1957. Google Scholar

[7] Waldron R A. Sense and Sense Development. London: Andre Deutsch, 1979. Google Scholar

[8] Lakoff G, Johnson M. Metaphor We Live by. Chicago: University of Chicago Press, 1980. Google Scholar

[9] Lakoff G, Espenson J, Schwartz A. Master Metaphor List. 2nd ed. Berkeley: University of California, 1991. Google Scholar

[10] Fass D. Met*: a method for discriminating metonymy and metaphor by computer. Comput Linguist, 1991, 17: 49--90. Google Scholar

[11] Martin J H. METABANK: A KNOWLEDGE-BASE OF METAPHORIC LANGUAGE CONVENTIONS. Comput Intelligence, 1994, 10: 134-149 CrossRef Google Scholar

[12] Krennmayr T, Steen G. VU Amsterdam metaphor corpus. In: Handbook of Linguistic Annotation. Berlin: Springer, 2017. 1053--1071. Google Scholar

[13] Zhu D X. Lecture on Chinese grammar. Beijing: Commercial Press, 1982. Google Scholar

[14] Yuan C F, Huang C N. Research on Chinese morphemes and Chinese word-formations based on the morpheme database. Chin Teach World, 1998, 2: 8--13. Google Scholar

[15] Kang S Y, Li Y, Sun D G, et al. The construction of Chinese systematic corpus and dictionary compilation. In: Proceedings of 2004 Symposium of Lexicography and Digitalization Seminar. Shanghai: Lexicographic Society of Shanghai, 2004. Google Scholar

[16] Li J F, Yang Y, Zhou C L. Corpus designing and constructing for metaphor computation. Mind Comput, 2007, 1: 142--146. Google Scholar

[17] Wang Z M. Chinese noun phrase metaphor recognition. Dissertation for Ph.D. Degree. Beijing: Peking University, 2007. Google Scholar

[18] Lu X, Wang B P Y. Towards a metaphor-annotated corpus of Mandarin Chinese. Lang Resources Evaluation, 2017, 51: 663-694 CrossRef Google Scholar

[19] Li J X, Li Y M. On the transparency of lexical meaning. Stud Lang Linguist, 2008, 28: 60--65. Google Scholar

[20] Fu H Q. The relationship between word meaning and morphemic meanings. Lexicogr Stud, 1981, 1: 98--110. Google Scholar

[21] Xu T Q. The core character and the semantic phraseology in Chinese. Linguist Res, 1997, 3: 2--16. Google Scholar

[22] Liu S X. Descriptive Lexicography in Chinese. Beijing: Commercial Press, 1990. Google Scholar

[23] Tian Y H, Liu Y. Lexical knowledge representation and sense prediction of Chinese unknown words. J Chin Inf Proc, 2016, 30: 26--34. Google Scholar

[24] Kang S C, Liu Y. Semantic word-formation based Chinese word similarity computing. J Chin Inf Proc, 2017, 31: 94--101. Google Scholar

[25] Lin Z, Liu Y. Implanting rational knowledge into distributed representation at morpheme level. In: Proceedings of the 33rd AAAI Conference on Artificial Intelligence, Honolulu, 2019. Google Scholar

[26] Goatly A. The Language of Metaphors. Abingdon: Routledge, 2011. Google Scholar

[27] Sweetser E. From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge: Cambridge University Press, 1991. Google Scholar

[28] Stern G. Meaning and Change of Meaning: With Reference to the English Language. Bloomington: Indiana University Press, 1931. Google Scholar

[29] Xu Y, Malt B C, Srinivasan M. Evolution of polysemous word senses from metaphorical mappings. In: Proceedings of the 38th Annual Meeting of the Cognitive Science Society, 2016. 2213--2218. Google Scholar

[30] Shutova E. Annotation of linguistic and conceptual metaphor. In: Handbook of linguistic annotation. Berlin: Springer, 2017. 1073--1100. Google Scholar

[31] Lakoff G, Turner M. More Than Cool Reason: A Field Guide to Poetic Metaphor. Chicago: University of Chicago Press, 1989. Google Scholar

[32] Deignan A. Metaphor and Corpus Linguistics. Amsterdam: John Benjamins Publishing, 2005. Google Scholar

  • Table 1   Examples of semantic word-formation in Chinese
    Example POS Word-formation pattern 1st-morpheme POS 2nd-morpheme POS 1st-sememe 2nd-sememe
    选材(Choose-talent) Verb Verb-object Verbal Nominal 选1_04_01 材1_05_04
    铁窗(Iron-window) Noun Modifier-head Nominal Nominal 铁1_07_01 窗1_01_01
  • Table 2   Examples of POS changes
    Word Definition POS change Metonymy direction
    领导$_2$(Lead-guide) Leader V$\rightarrow$N Act$\rightarrow$Subject
    收入$_2$(Receive-in) Income V$\rightarrow$N Act$\rightarrow$Object
    摆渡$_3$(Take-ferry) Ferry V$\rightarrow$N Act$\rightarrow$Instrument
    撰述$_2$(Write-tell) Written work V$\rightarrow$N Act$\rightarrow$Result
    整齐$_2$(Orderly-neat) To make neat A$\rightarrow$V Attribute$\rightarrow$Cause
    无辜$_3$(No-sin) The innocent A$\rightarrow$N Attribute$\rightarrow$Subject
    新潮$_2$(New-tide) Modern N$\rightarrow$A Object$\rightarrow$Attribute
    补益$_2$(Advantage-benefit) To benefit N$\rightarrow$V Object$\rightarrow$Act
  • Table 3   Examples of other types
    Word Definition Metonymy direction
    花卉$_2$(Flower-flower) Chinese painting with flowers Content$\rightarrow$Artwork
    文书$_2$(Document-book) Secretary Work$\rightarrow$Worker
    官府$_2$(Officer-mansion) Feudal bureaucrat Working place$\rightarrow$Worker
    割席 (Split-mat) Break off a relation An act in a story$\rightarrow$The incident ensuing from the act
  • Table 4   Examples of synonyms with literal meanings for Chinese non-literal words
    Example Definition Type Sub-type Source Target Computational Literal meaning
    bearer of bearer
    铁窗(Iron-window) Jail Metonymy 2 Bn04 Dm02 监狱(jail-jail) [狱1\_02\_01,~监2\_02\_02]
    全豹(Whole-leopard) The whole thing Metaphor 1 Bi02 Da21 全貌(Whole-aspect) [貌1\_03\_02,~全1\_05\_03]
  • Table 5   The distribution of metaphor categories in Chinese words
    Metaphorical type Number Percentage (%) Example Literal meaning Non-literal meaning
    #1 1160 58.1 主将$_2$ (Main-general) [将1\_02\_01,~主1\_13\_06] [力1\_05\_02,~主1\_13\_06]
    #2 243 12.2 失足$_2$ (Lose-foot) [失1\_07\_02,~足1\_04\_01] [堕1\_01\_01,~落1\_12\_04]
    #3 73 3.7 沉浸 (Sink-immerse) [沉1\_06\_01,~浸1\_03\_01] [处2\_06\_03,~于1\_02\_01]
    #4 90 4.5 复活 (Again-live) [活1\_06\_01,~复3\_02\_01] [兴2\_07\_01,~复3\_02\_01]
    #5 109 5.5 保守$_2$ (Keep-guard) [保1\_07\_02,~守1\_05\_01] [守1\_05\_03,~旧1\_05\_01]
    #1, #4 6 0.3 葛藤 (Kudzu-vine) [藤1\_02\_01,~葛2\_02\_01] [纠1\_02\_01,~纷1\_02\_02]
    #1, #5 234 11.7 上供$_2$ (Offer-offerings) [上2\_14\_03,~供1\_02\_02] [行4\_13\_07,~贿1\_02\_02]
    #2, #5 39 2.0 插手$_2$ (Insert-hand) [插1\_02\_02,~手1\_07\_01] [介1\_04\_01,~入1\_05\_01]
    #3, #5 11 0.6 气焰 (Gas-flame) [气1\_14\_06,~焰1\_01\_01] [势1\_06\_01,~焰1\_01\_01]
    #4, #5 30 1.5 羽翼(Feather-wing) [羽1\_04\_02,~翼1\_07\_01] [手1\_07\_07,~助1\_01\_01]
    #1, #4, #5 2 0.1 鸡肋 (Chicken-rib) [肋1\_01\_01,~鸡1\_02\_01] [事1\_06\_01,~琐1\_02\_01]
  • Table 6   The distribution of metonymy categories in Chinese words
    Metonymic type Number Percentage (%) Example Literal meaning Non-literal meaning
    #1 750 49.1 编辑$_2$ (Edit-edit) [编1\_09\_03,~辑1\_02\_01] [者1\_06\_01,~编1\_09\_03]
    #2 190 12.4 铁窗 (Iron-window) [窗1\_01\_01,~铁1\_07\_01] [狱1\_02\_01,~监2\_02\_02]
    #3 587 38.4 官府 (Officer-mansion) [府1\_06\_01,~官1\_04\_01] [官1\_04\_01,~吏1\_03\_02]
  • Table 7   Examples of source and target domains with dominant distribution in metaphor
    Dominant source domain Number Example Dominant target domain Number Example
    Bk08 (Limb) 26 巨擘、手心$_2$、肘腋 Da21 (Situation) 29 地狱$_2$、平台$_4$、底牌$_2$
    Dh01 (Devil) 24 凶神、天仙、夜叉 Da20 (Trend) 26 春潮、暗流$_2$、死棋
    Hb02 (March and battle) 16 挑战$_2$、收兵$_2$、攻关 Ig01 (Begin and end) 21 萌发$_2$、起步$_2$、上马
  • Table 8   Examples of source and target domains with dominant distribution in metonymy
    Dominant source domain Number Example Dominant target domain Number Example
    Bk02 (Head or face) 13 头脸、头面、秃头$_4$ Dj08 (Fund or cost) 27 积蓄$_2$、茶钱$_2$、开销$_2$
    Hc09 (Preside or rule) 12 指挥$_2$、领队$_2$、主席 Ae01 (Worker or clerk) 26 剧务$_2$、总务$_2$、文书$_2$
    Dao1 (Things) 12 剧务$_2$、常务、机密 Af10 (Leader) 20 总裁、主席、领导$_2$
  • Table 9   The distribution of distances between source and target domains in metaphor and metonymy
    Location between Number of Percentage (%) Number of Percentage (%)
    source and target metaphoric words metonymic words
    In the same hang of the same ciqun 0 0 0 0
    In different hangs of the same ciqun 90 4.5 84 5.5
    In different ciquns of the same xiaolei 14 0.7 13 0.9
    In different xiaoleis of the same zhonglei 108 5.4 66 4.3
    In different zhongleis of the same dalei 420 21.0 170 11.1
    In different daleis 1365 68.4 444 (1194) 29.1 (78.3)
  • Table 10   The comparison of word similarity calculation results between the original method and method in this paper
    Word1 Non-literal Word2 Non-literal Original Ours
    meaning of word1 meaning of word2
    司机(Steer-machine) [手1\_07\_07,~车1\_07\_01] 开车(Drive-car) [开2\_18\_06,~车1\_07\_01] 0.527 0.083
    司机(Steer-machine) [手1\_07\_07,~车1\_07\_01] 乘客(Ride-guest) [客1\_10\_05,~乘4\_04\_01] 0.304 0.642
    铁窗(Iron-window) [狱1\_02\_01,~监2\_02\_02] 纱窗(Yarn-window) [窗1\_01\_01,~纱1\_04\_03] 0.967 0.780
    铁窗(Iron-window) [狱1\_02\_01,~监2\_02\_02] 牢房(Jail-house) [房1\_07\_02,~牢1\_05\_03] 0.374 0.910
    脾胃(Spleen-stomach) [志1\_03\_01,~趣1\_04\_03] 肠胃(Intestine-stomach) [肠1\_03\_01,~胃1\_02\_01] 1.000 0.282
    脾胃(Spleen-stomach) [志1\_03\_01,~趣1\_04\_03] 兴趣(Mood-interest) [兴1\_01\_01,~趣1\_04\_03] 0.542 0.917
    全豹(Whole-leopard) [貌1\_03\_02,~全1\_05\_01] 老虎(Old-tiger) [虎1\_04\_01,~老1\_17\_16] 0.895 0.336
    全豹(Whole-leopard) [貌1\_03\_02,~全1\_05\_01] 全局(Whole-situation) [局1\_10\_05,~全1\_05\_03] 0.426 0.648

Contact and support