: 117 2, 3, 4, 5,, 6, 2,,, ( ) ( ) ( ) ( ) [10 ] ( ),,,,,,,, 1 1,

Size: px

Start display at page:

Download ": 117 2, 3, 4, 5,, 6, 2,,, ( ) ( ) ( ) ( ) [10 ] ( ),,,,,,,, 1 1,"

稍诸选能
5 years ago
Views:

1 J OU RNAL OF CH IN ESE IN FORMA TION PROCESSIN G Vol. 22, No. 1 Jan., 2008 : (2008) ,, (, ) :,: 4,100,, : ; ; ; ; ; : TP391 : A Construction and Analysis of Emotional Corpus XU Lin2hong, L IN Hong2fei, ZHAO Jing (Department of Computer Science and Engineering, Dalian University of Technology, Dalian, Liaoning , China) Abstract : This paper introduced some experiences on constructing emotional corpus, and discussed several basic questions which included the tagging criterion, tagging set, tagging tools and quality monitoring. There were about sentences in the corpus. Moreover based on these, statistical data about emotional distribution and rules of emotional transference were available, and characters and applications of corpus were analyzed, so emotional corpus provide support for text affective computing. Key words : computer application ; Chinese information processing ; emotional corpus ; text coding initiative ; consistency checking ; emotional transference 1,,,,,,, Brown 20 80, [1 ] [ 2 ] [3 ] [4 ],, Pang [5 ], Whissell [6 ],Berardinelli [7 ], [8 ] [9 ],,,,,, : : : ( , ) ; 863 (2006AA01Z151) ; : (1979 ),,, ; (1962 ),,,,, ; (1961 ),,,,

2 : 117 2, 3, 4, 5,, 6, 2,,, ( ) ( ) ( ) ( ) [10 ] ( ),,,,,,,, 1 1, ,, ;,, [ 11 ],,,,, 3. 1,, persons = ( persona1, persona2,,,,, : DocumentM odel = ( title, aut hor, st y le, source, persons, sentences, key note) (1) SentenceM odel = ( ori gi n, sender, [ accepter ], [ rhetoric ], emotions, [ key w ords ]) (2), accepter rhetoric key w ords, 2 persons sentences emotions keywords,,, personai ), sender accepter persons persons

118 2008 2 (documentmodel) ( sentencemodel) title author,, st yle source persons 1 2 i sentences sentencemodel key note o h e i m f d s ori gin sender i accepter i rhetoric emotions o h e p r b l k c

3 (documentmodel) ( sentencemodel) title author,, st yle source persons 1 2 i sentences sentencemodel key note o h e i m f d s ori gin sender i accepter i rhetoric emotions o h e p r b l k c i s w g m u f x t d a j y q key w ords 1 2 i,,,,,, sentences, key w ords, key w ords, [12 ],, [13 ],,, keynote 1 emotions TEI Leech, (1993) ; ;,, TEI ( Text Encoding Initiative), TEI TEI ( ACL, Association for Comp utational Linguistics ) ( ALL C, Association for Literary and LinguisticComp uting) ( AC H, A ssociatio n for Co mp uters and

4 : 119 Humanities) ( The British National Corp us) TEI TEI, ( Header),, TEI (Start tag) ( End tag), p / p,, TEI 2 2 document / document, header / header, body / body p / p, sect / sect title orgin, emotions Leech,, xml, ,,, :,, 2, 1 3,,,,, 23,,,,,,

5 , 3,,,,, [14 ], ,,,,,,,,,3 3 Neg Si i 0 1 larsame Ei, Ej i j ( ) 0 wordsame Si, Sj i j 0 personsame Si i - 1, i i emotionsame Ei i - 1, i i + 1 0,, wordconsistency ( i, j) = words ame ( S i, S j ) lars ame ( Ei, Ej ) Neg ( S i) Neg ( S j ) (3) S i S j i j, Ei E j i j w ordconsistency,,,, 1,,, (4) : contex t Consistency ( i) = persons ame ( S i) emotions ame ( Ei) (4),,

6 : 121,,,, , ,39 488, ,, ,, 2 000, ,, ( ) (5) : = trans f er ( Ea, Eb) n m i i = 1 j = 1 (equal ( Ej, Ea) equal ( Ej+1, Eb) ) Ta + Tb (5) t rans f er( Ea, Eb) a b, n, mi i, Ti i Ei E j, equal ( Ei, Ej ) 1, 0 a b, 4 23, 23,, t rans f er ( Ea, Eb) ( ), ,,,

7 ,,, 4 6, ,39 488,, ,, : [1 ]. [J ]., 1996, (3) :229. [2 ] http :/ / www. sinica. edu. tw/ SinicaCorpus/ [DB/ OL ]. [3 ],,. [J ]., 1997, (2) : [4 ] http :/ / www. icl. pku. edu. cn/ icl _ group s/ corpustagging. asp [DB/ OL ]. [5 ] http :/ / www. cs. cornell. edu/ People/ pabo/ movie2 review2data/ [DB/ OL ]. [6 ] Theologos Athanaselis, Stelios Bakamidis, and Ioannis Dologlou. Recognizing Verbal Content of Emotionally Colored Speech [ A ]. European Signal Processing Conference[ C] [7 ] http :/ / www. reelviews. net/ [DB/ OL ]. [8 ] http :/ / epinions. com/ [DB/ OL ]. [9 ] Hongwu Yang, Helen M. Meng, Zhiyong Wu and Lianhong Cai. Modeling the Global Acoustic Correlates of Expressivity for Chinese Text2to2Speech Synthesis [ A ]. IEEE / ACL 2006 Workshop on Spoken Language Technology [ C ]. Aruba, [10 ]. [J ]. 1999, (1) : [11 ]. [J ].,1994,8 (3) : [12 ],. [J ].,2007, 44 (S2) : [13 ],,. [J ].,2007,21 (1) : [14 ] Christopher D. Manning, Hinrich Schutze. [ M ]. :, ( 66 ) [8 ] Berger, A. and Lafferty, J. D.. Information Retrieval as Statistical Translation [ A ]. In : proceedings of the 1999 ACM SIGIR Conference on Research and Development in Information Retrieval [ C]. 1999, [9 ] Andreas Hotho, Steffen Staab, and Gerd Stumme. Ontologies improve text document clustering [ A ]. In : Proc. of the ICDM 03, The 2003 IEEE International Conference on Data Mining, [10 ] Zhou, X., Zhang, X., and Hu, X., The Dragon Toolkit, Data Mining & Bioinformatics Lab, ischool at Drexel University, http :/ / www. ischool. drexel. edu/ dmbio/ dragontool [ CP/ OL ]. [11 ] Zhou X., Hu X., Zhang X.. Using Concept2Based Indexing to Improve Language Modeling Approach to Genomic IR [ A ]. ECIR 2006 [ C ]. LNCS 3936, [12 ] J. C. Bezdek. Pattern Recognition with Fuzzy Objective Function Algorithms [ M ]. Plenum Press, New York [13 ] Dempster, A. P., Laird, N. M., and Rubin, D. B., Maximum likelihood from incomplete data via the EM algorithm [ J ]. Journal of the Royal Statistical Society, 1977, 39 : [14 ] http :/ / ir. ohsu. edu/ genomics/ data/ [DB/ OL ]. [ 15 ] Hersh W, et al. TREC 2004 Genomics Track Overview [ A ]. the thirteenth Text Retrieval Conference [ C] [ 16 ] Hersh W, et al. TREC 2005 Genomics Track Overview [ A ]. the fourteenth Text Retrieval Conference [ C] [17 ],,. [J ].,2005,19 (3) :

,,, () 20 80,,,,, ;,, ;,, ;,,,,,,,,, [1 ], :,,,,2 2,,, () (),,,,:,,,,:,,,, :, [2 ] :,,,,,,, : AN NA,,,,,, ( ),:,,: ( F) = (A1 + A2 + A3 + An -

,,, () 20 80,,,,, ;,, ;,, ;,,,,,,,,, [1 ], :,,,,2 2,,, () (),,,,:,,,,:,,,, :, [2 ] :,,,,,,, : AN NA,,,,,, ( ),:,,: ( F) = (A1 + A2 + A3 + An - 23 5 2009 9 J OU RNAL OF CH IN ESE IN FORMA TION PROCESSIN G Vol. 23, No. 5 Sep., 2009 : 100320077 (2009) 0520009210, (,) :,, ;,,,, ;,, : ;; ;;; : TP391 : A A Semantic Construction Model bet ween Adjectives