| { |
| "paper_id": "O03-1014", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T08:01:56.880557Z" |
| }, |
| "title": "Extraction and Analysis of Research Topics Based on NLP Technologies", |
| "authors": [ |
| { |
| "first": "Sung-Chen", |
| "middle": [], |
| "last": "Lin", |
| "suffix": "", |
| "affiliation": { |
| "laboratory": "", |
| "institution": "Shih-Hsin University", |
| "location": {} |
| }, |
| "email": "scl@cc.shu.edu.tw" |
| } |
| ], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "", |
| "pdf_parse": { |
| "paper_id": "O03-1014", |
| "_pdf_hash": "", |
| "abstract": [], |
| "body_text": [ |
| { |
| "text": "\u8cc7\u8a0a\u6aa2\u7d22\u7814\u7a76\u8457\u91cd\u7684\u554f\u984c\u662f\u4eba\u8207\u8cc7\u8a0a\u4e4b\u9593\u7684\u4ecb\u9762\uff0c\u8fd1\u4f86\u7684\u7814\u7a76\u8da8\u52e2\u6ce8\u91cd\u65bc\u4f7f \u7528\u8005\u6240\u5177\u6709\u7684\u80cc\u666f\u77e5\u8b58\u3001\u5728\u6aa2\u7d22\u904e\u7a0b\u4e2d\u5c0d\u554f\u984c\u7684\u8a8d\u77e5 [Wilson, 1999] \u53ca\u8cc7\u6599\u7684\u5afb\u719f \u7a0b\u5ea6(material mastery) [Bishop, 1999] [Covi, 1999] [Harter, 1992] [Wayne, 2000] \u3002\u76ee\u524d\u7814\u7a76\u4eba\u54e1\u8a8d\u70ba\u300e\u53e2\u805a\u5047\u8aaa\u300f(cluster hypothesis) \u53ef \u4ee5 \u9069 \u7528 \u65bc \u89e3 \u6c7a \u9019 \u500b \u554f \u984c [Yang, Pierce and Carbonell, 1998 ] [Hatzivassiloglou, Gravano and Maganti, 2000] [Yang, Pierce and Carbonell, 1998 ]\u3002\u4f46\u662f\u5b78\u8853\u8ad6\u6587\u96d6\u7136\u6709\u6240\u8b02\u300e\u8cc7\u8a0a\u6d41\u884c\u300f(information epidemic) [Tabah, 1996] \u7684\u8aaa\u6cd5\uff0c\u4e5f\u5c31\u662f\u5728\u67d0\u4e00\u9805\u65b0\u7684\u7406\u8ad6\u3001\u7814\u7a76\u65b9\u6cd5\u6216\u6280\u8853\u63d0\u51fa\u5f8c\uff0c\u5982\u679c\u5f97\u5230\u5f88\u5927\u7684\u6210 \u529f\uff0c\u5c07\u53ef\u4ee5\u5438\u5f15\u8a31\u591a\u7814\u7a76\u4eba\u54e1\u6295\u5165\u5f8c\u7e8c\u7684\u7814\u7a76\u4e2d\uff0c\u9020\u6210\u4e00\u80a1\u76f8\u95dc\u7814\u7a76\u767c\u8868\u7684\u98a8\u6f6e\uff0c \u7136\u800c\u5728\u5be6\u8b49\u7814\u7a76\u4e2d\u537b\u767c\u73fe\u6b64\u4e00\u73fe\u8c61\u96d6\u7136\u5b58\u5728\u4f46\u4e26\u4e0d\u5e38\u898b [Tabah, 1996] (1) ", |
| "cite_spans": [ |
| { |
| "start": 58, |
| "end": 72, |
| "text": "[Wilson, 1999]", |
| "ref_id": "BIBREF13" |
| }, |
| { |
| "start": 101, |
| "end": 115, |
| "text": "[Bishop, 1999]", |
| "ref_id": null |
| }, |
| { |
| "start": 116, |
| "end": 128, |
| "text": "[Covi, 1999]", |
| "ref_id": null |
| }, |
| { |
| "start": 129, |
| "end": 143, |
| "text": "[Harter, 1992]", |
| "ref_id": "BIBREF5" |
| }, |
| { |
| "start": 144, |
| "end": 157, |
| "text": "[Wayne, 2000]", |
| "ref_id": "BIBREF12" |
| }, |
| { |
| "start": 216, |
| "end": 249, |
| "text": "[Yang, Pierce and Carbonell, 1998", |
| "ref_id": "BIBREF14" |
| }, |
| { |
| "start": 252, |
| "end": 297, |
| "text": "[Hatzivassiloglou, Gravano and Maganti, 2000]", |
| "ref_id": "BIBREF6" |
| }, |
| { |
| "start": 298, |
| "end": 331, |
| "text": "[Yang, Pierce and Carbonell, 1998", |
| "ref_id": "BIBREF14" |
| }, |
| { |
| "start": 374, |
| "end": 387, |
| "text": "[Tabah, 1996]", |
| "ref_id": null |
| }, |
| { |
| "start": 486, |
| "end": 499, |
| "text": "[Tabah, 1996]", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "\u4e00\u3001\u7dd2\u8ad6", |
| "sec_num": null |
| }, |
| { |
| "text": "\u5728\u5f0f(1)\uff0cm S \u548c \u03c3 S \u5206\u5225\u4ee3\u8868\u5b57\u4e32 S \u5728\u51fa\u73fe\u8ad6\u6587\u4e2d\u7684\u5e73\u5747\u983b\u6b21\u548c\u6a19\u6e96\u5dee\u3002\u7576\u5b57\u4e32 S \u7684\u5e73\u5747\u983b\u6b21\u8d85\u904e\u67d0\u4e00\u95be\u503c\u6642\uff0c\u8868\u793a\u6b64\u5b57\u4e32\u6975\u6709\u53ef\u80fd\u5728\u8a31\u591a\u8ad6\u6587\u4e2d\u51fa\u73fe\u591a\u6b21\uff0c\u662f\u9019 \u4e9b\u8ad6\u6587\u7684\u95dc\u9375\u8a5e\u8a9e\uff0c\u61c9\u8a72\u88ab\u9078\u53d6\u51fa\u4f86\u3002\u6216\u662f\u96d6\u7136\u5b57\u4e32 S \u5728\u8ad6\u6587\u7684\u5e73\u5747\u983b\u6b21\u8f03\u4f4e\uff0c \u4f46\u5728\u67d0\u4e9b\u8ad6\u6587\u4e2d\u51fa\u73fe\u591a\u6b21\uff0c\u662f\u9019\u4e9b\u8ad6\u6587\u7684\u95dc\u9375\u8a5e\u8a9e\uff0c\u4e5f\u9700\u8981\u88ab\u9078\u53d6\u51fa\u4f86\uff0c\u6b64\u6642\u5b57 \u4e32 S \u6703\u6709\u4e00\u500b\u8f03\u5927\u7684\u6a19\u6e96\u5dee \u03c3 S \u3002\u56e0\u6b64\uff0c\u6211\u5011\u53ef\u4ee5\u5229\u7528\u5b57\u4e32\u5728\u51fa\u73fe\u8ad6\u6587\u4e2d\u7684\u5e73\u5747\u983b \u6b21\u548c\u6a19\u6e96\u5dee\u7684\u7e3d\u548c R S \u4ee3\u8868\u5b57\u4e32\u5c0d\u51fa\u73fe\u8ad6\u6587\u7684\u91cd\u8981\u7a0b\u5ea6\uff0cR S \u503c\u6108\u9ad8\u7684\u5b57\u4e32\u5c0d\u51fa\u73fe\u8ad6 \u6587\u6108\u91cd\u8981\u3002 \u5b57\u4e32\u524d\u5f8c\u63a5\u5b57\u7684\u8907\u96dc\u5ea6\u5247\u53ef\u4ee5\u5224\u65b7\u662f\u5426\u662f\u4e00\u500b\u5b8c\u6574\u7684\u8a5e\u8a9e\u6216\u662f\u5176\u4ed6\u8a5e\u8a9e\u7684\u90e8 \u5206\uff0c\u5b57\u4e32 S \u7684\u524d\u5f8c\u63a5\u5b57\u8907\u96dc\u5ea6 C 1S \u548c C 2S \u5206\u5225\u5982\u5f0f(2a)\u548c(2b)\u6240\u793a ) log( 1 S aS a S aS def S F F F F C \u2211 \u2212 = (2a) ) log( 2 S Sb b S Sb def S F F F F C \u2211 \u2212 = (2b) \u5f0f(2a)\u548c(2b)\u4e2d\uff0ca \u548c b \u4ee3\u8868\u5b57\u4e32 S \u5728\u8ad6\u6587\u8cc7\u6599\u4e2d\u4efb\u4e00\u500b\u53ef\u80fd\u7684\u524d\u63a5\u5b57\u548c\u5f8c\u63a5 \u5b57\uff0cF S \u3001F aS \u548c F Sb \u5206\u5225\u662f\u5b57\u4e32 S\u3001aS \u548c Sb \u7684\u51fa\u73fe\u7e3d\u983b\u6b21\u3002\u4ee5\u5f0f(2a)\u524d\u63a5\u5b57\u7684\u60c5\u5f62 \u4f86\u770b\uff0c\u82e5\u662f\u5b57\u4e32 S \u6709\u6108\u591a\u7a2e\u985e\u7684\u524d\u63a5\u5b57\uff0c\u800c\u4e14\u6bcf\u4e00\u7a2e\u524d\u63a5\u5b57\u51fa\u73fe\u7684\u6b21\u6578\u8d8a\u63a5\u8fd1\u6642\uff0c C 1S \u7684\u503c\u6108\u5927\uff0c\u53cd\u4e4b\uff0c\u7576\u5b57\u4e32\u524d\u53ea\u6709\u4e00\u7a2e\u524d\u63a5\u5b57\u6642\uff0cC 1S \u7684\u503c\u7b49\u65bc 0\uff0c\u6216\u662f\u6709\u4e00\u500b\u524d \u63a5\u5b57\u51fa\u73fe\u7684\u6a5f\u6703\u8f03\u5176\u4ed6\u5927\u975e\u5e38\u591a\u6642\uff0c\u5247 C 1S \u7684\u503c\u63a5\u8fd1\u65bc", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "\u4e00\u3001\u7dd2\u8ad6", |
| "sec_num": null |
| }, |
| { |
| "text": "EQUATION", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [ |
| { |
| "start": 0, |
| "end": 8, |
| "text": "EQUATION", |
| "ref_id": "EQREF", |
| "raw_str": "\u8a9e A \u8207 B \u9593\u7684\u76f8\u95dc\u7a0b\u5ea6\u4f30\u7b97\u65b9\u5f0f\u3002 ] ,..., , [ ' , , 2 , 1 A N A A def A f f f v = r (3) B A B A def v v v v B A R r r \u22c5 = ) , (", |
| "eq_num": "(4)" |
| } |
| ], |
| "section": "\u4e00\u3001\u7dd2\u8ad6", |
| "sec_num": null |
| }, |
| { |
| "text": "r r [Deerwester, et. al., 1990] ", |
| "cite_spans": [ |
| { |
| "start": 4, |
| "end": 31, |
| "text": "[Deerwester, et. al., 1990]", |
| "ref_id": "BIBREF4" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "\u4e00\u3001\u7dd2\u8ad6", |
| "sec_num": null |
| }, |
| { |
| "text": "\u5f0f(3)\u4e2d\uff0cf i,A \u4ee3\u8868\u8a5e\u8a9e A \u5728\u7b2c i \u7bc7\u8ad6\u6587\u8cc7\u6599\u4e2d\u51fa\u73fe\u7684\u983b\u6b21\u3002\u5f0f(4)\u4e2d\uff0c\u5206\u5b50\u90e8\u5206 \u662f\u8a5e\u8a9e A \u548c B \u7684\u7279\u5fb5\u5411\u91cf \u548c v A v r B r \u5167\u7a4d(inner product)\u7684\u503c\uff0c\u5206\u6bcd\u90e8\u5206\u5247\u662f\u5169\u500b\u7279\u5fb5 \u5411\u91cf\u9577\u5ea6 A v r \u548c B v", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "\u4e00\u3001\u7dd2\u8ad6", |
| "sec_num": null |
| }, |
| { |
| "text": "\uff0c\u5c07\u77e9\u9663 M \u5206\u89e3\u6210\u4e09\u500b\u77e9\u9663\uff0cT o \u3001S o \u548c D o \uff0c\u4f7f\u5f97 '\u3002\u6b64\u8655 T o o o D S T M = o \u548c D o \u70ba M \u7684\u5de6\u3001\u53f3\u5947\u7570\u5411\u91cf(singular vectors)\u6240\u5f62\u6210\u7684\u77e9\u9663\uff0c \u5176\u5927\u5c0f\u5206\u5225\u70ba t\u00d7r \u548c d\u00d7 r\uff0ct \u548c d \u5206\u5225\u70ba\u8a5e\u8a9e\u548c\u7279\u5fb5\u7684\u6578\u76ee\uff0cr \u5247\u70ba\u77e9\u9663 M \u7684\u79e9 (rank)\uff0c\u800c S o \u70ba\u4e00\u500b\u5927\u5c0f\u70ba r\u00d7r \u7684\u5c0d\u89d2\u7dda\u77e9\u9663(diagonal matrix)\uff0c\u5176\u5c0d\u89d2\u7dda\u4e0a\u7684\u503c\u70ba M \u7684\u5947\u7570\u503c(singular values)\uff0c\u4e14\u4f9d\u64da\u905e\u6e1b\u7684\u65b9\u5f0f\u6392\u5217\u3002\u82e5\u6211\u5011\u5e0c\u671b\u53d6\u5f97\u4e00\u500b\u79e9\u70ba k \u7684\u77e9\u9663 M \uff0ck<=r\uff0c\u4e26\u4f7f\u5f97 M \u8207 M \u7684\u6700\u5c0f\u5e73\u65b9\u5dee(least square error)\u6700\u63a5\u8fd1\uff0c\u53ef\u4ee5\u53d6 S o \u5c0d\u89d2\u7dda\u4e0a\u7684\u524d k \u500b\u5947\u7570\u503c\uff0c\u7522\u751f\u4e00\u500b\u5927\u5c0f\u70ba k\u00d7k \u7684\u65b0\u77e9\u9663 S\uff0c\u540c\u6642 T o \u548c D o \u4e5f\u5206 \u5225\u53d6\u524d k \u500b\u884c\u5411\u91cf(column vectors)\uff0c\u5f62\u6210\u77e9\u9663 T \u548c D\uff0c\u5927\u5c0f\u5206\u5225\u70ba t\u00d7k \u548c d\u00d7 k\u3002\u77e9 \u9663 M \u4fbf\u53ef\u7531 \u8a08\u7b97\u5f97\u5230\u3002\u5728\u4f7f\u7528 LSI \u6280\u8853\u7684\u6aa2\u7d22\u904e\u7a0b\uff0c\u7576\u9032\u884c\u8a5e\u8a9e\u7684\u76f8\u95dc \u7a0b\u5ea6\u4f30\u7b97\u6642\uff0c\u4ee5 ' TSD M = 'M M \u4f86\u4f30\u7b97\u539f\u5148\u4ee5 MM'\u8a08\u7b97\u5169\u5169\u8a5e\u8a9e\u7279\u5fb5\u5411\u91cf\u9593\u7684\u5167\u7a4d\u503c\uff0c\u5982\u5f0f (5)\u6240\u8868\u793a\uff0c 'M ' MM ' ' ' ' ' ' )' ' ( ' 2 T TS T TSS T DS TSD TSD TSD M = = = = \u2248 (5) \u5728\u5f0f(5)\u4e2d\uff0c\u7531\u65bc\u77e9\u9663 D \u4e2d\u7684\u884c\u5411\u91cf\u5f7c\u6b64\u4e92\u70ba\u55ae\u4f4d\u6b63\u4ea4(othonormal)\uff0cDD'=I\uff0c", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "\u4e00\u3001\u7dd2\u8ad6", |
| "sec_num": null |
| }, |
| { |
| "text": "\u800c\u4e14 S \u70ba\u5c0d\u89d2\u7dda\u77e9\u9663\uff0cS'= S\uff0c\u6240\u4ee5 '\u3002\u5229\u7528 SVD \u53d6\u5f97\u96b1\u542b\u8a9e\u7fa9\u7d50\u69cb(latent semantic structure)\u7684\u7279\u6027\uff0c\u4f7f\u5f97\u539f\u5148\u56e0\u70ba\u5171\u73fe\u95dc\u4fc2\u8f03\u5f31\u6216\u662f\u4e0d\u5b58\u5728\uff0c\u800c\u76f8\u95dc\u7a0b\u5ea6\u8f03 \u4f30\u7b97\u5f97\u5f88\u5c0f\u7684\u5169\u500b\u76f8\u95dc\u8a5e\u8a9e\uff0c\u53ef\u4ee5\u7372\u5f97\u8f03\u5927\u7684\u4f30\u7b97\u503c [Deerwester, et. al., 1990] \u3002 [Deerwester, et. al., 1990 ]\uff0c\u5982\u5f0f(6)\u8a08\u7b97\u53e2\u96c6\u03a7\u5c0d\u6240\u6709\u8ad6\u6587 ", |
| "cite_spans": [ |
| { |
| "start": 118, |
| "end": 145, |
| "text": "[Deerwester, et. al., 1990]", |
| "ref_id": "BIBREF4" |
| }, |
| { |
| "start": 148, |
| "end": 174, |
| "text": "[Deerwester, et. al., 1990", |
| "ref_id": "BIBREF4" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "\u4e00\u3001\u7dd2\u8ad6", |
| "sec_num": null |
| }, |
| { |
| "text": "' 2 T TS M M = \u5176\u6b21\uff0c\u9032\u884c cliques \u53e2\u96c6\u6f14\u7b97\u6cd5\u5f8c\uff0c\u6211\u5011\u5c0d\u65bc\u6240\u5f97\u5230\u7684\u7d50\u679c\u4f9d\u64da\u5b83\u5011\u6210\u54e1\u9593\u91cd \u758a\u7684\u60c5\u5f62\u518d\u6b21\u9032\u884c\u53e2\u96c6\u3002\u5047\u8a2d\u5169\u500b\u53e2\u96c6\u4e4b\u9593\u6709\u4e09\u500b\u4ee5\u4e0a\u7684\u6210\u54e1\u662f\u76f8\u540c\u7684\uff0c\u800c\u4e14\u5176 \u9918\u7684\u6210\u54e1\u9593\u96d6\u7136\u6c92\u6709\u5f88\u5f37\u7684\u8a5e\u8a9e\u5171\u73fe\u95dc\u4fc2\uff0c\u4f46\u662f\u4e5f\u66fe\u5728\u67d0\u4e9b\u8ad6\u6587\u8cc7\u6599\u4e2d\u4e00\u8d77\u51fa \u73fe\uff0c\u6211\u5011\u5373\u5c07\u9019\u5169\u500b\u53e2\u96c6\u7684\u8a5e\u8a9e\u96c6\u5408\u9032\u884c\u806f\u96c6\uff0c\u7522\u751f\u65b0\u53e2\u96c6\u3002\u5982\u5716\u4e8c\u6240\u793a\uff0c \u5728 A\u3001 B\u3001C\u3001D\u3001E \u548c F \u516d\u500b\u76f8\u95dc\u8a5e\u8a9e\u4e2d\uff0c\u4f9d\u64da\u5b83\u5011\u7684\u5171\u73fe\u95dc\u4fc2\u9032\u884c cliques \u53e2\u96c6\uff0c\u53e2\u96c6 \u6210\u03a7 1 \u548c\u03a7 2 \u5169\u500b\u8a5e\u8a9e\u96c6\u5408\u3002\u5728\u9019\u5169\u500b\u53e2\u96c6\u9593\uff0c\u6709\u4e09\u500b\u8a5e\u8a9e A\u3001B \u548c C \u662f\u76f8\u540c\uff0c\u800c\u4e14 A B C D E F A B C D A B C E F \u03a7 1 \u03a7 2 \u5716\u4e8c \u5c07\u03a7 1 \u548c\u03a7 2 \u5169\u500b\u5177\u6709\u76f8\u540c\u6210\u54e1\u7684\u53e2\u96c6\u9032\u884c\u5408\u4f75\u7684\u793a\u610f\u5716 \u6700\u5f8c\u7d93\u904e\u4e0a\u8ff0\u7684\u53e2\u96c6\u8655\u7406\u5f8c\uff0c\u53ef\u4ee5\u5f97\u5230\u4e00\u4e9b\u4ee3\u8868\u9818\u57df\u4e2d\u91cd\u8981\u7814\u7a76\u4e3b\u984c\u7684\u8a5e\u8a9e \u53e2\u96c6\u3002\u5728\u5206\u6790\u7814\u7a76\u4e3b\u984c\u6642\uff0c\u6211\u5011\u8a08\u7b97\u6bcf\u4e00\u53e2\u96c6\u8207\u8ad6\u6587\u9593\u7684\u76f8\u95dc\u7a0b\u5ea6\uff0c\u8a08\u7b97\u65b9\u5f0f\u4f9d \u64da\u70ba LSI \u7684\u76f8\u95dc\u4f30\u8a08\u65b9\u5f0f", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "\u4e00\u3001\u7dd2\u8ad6", |
| "sec_num": null |
| }, |
| { |
| "text": "EQUATION", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [ |
| { |
| "start": 0, |
| "end": 8, |
| "text": "EQUATION", |
| "ref_id": "EQREF", |
| "raw_str": "\u7684\u76f8\u95dc\u7a0b\u5ea6\u3002 ' TSD R \u03c7 = \u03a7 (6) \u5f0f(6)\u4e2d\uff0c\u03c7\u70ba\u4e00\u500b\u884c\u5411\u91cf\uff0c ] , ,", |
| "eq_num": ", [ ' 2" |
| } |
| ], |
| "section": "\u4e00\u3001\u7dd2\u8ad6", |
| "sec_num": null |
| }, |
| { |
| "text": "acquisition, explanation, generalization, learning", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "initial, min, taiwanese, \u53f0\u8a9e, \u53f0\u7063, \u8cc7\u6599\u5eab", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "hidden markov, maximum, robust speech recognition, speech recognition", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "document, text categorization, \u5206\u985e, \u6587\u4ef6\u5206\u985e, \u6587\u4ef6\u81ea\u52d5, \u95dc\u9375\u8a5e", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "speech, synthesis, \u6587\u53e5\u7ffb\u8a9e\u97f3, \u5408\u6210, \u7cfb\u7d71, \u97f3\u7bc0, \u570b\u8a9e, \u9023\u97f3, \u8a9e\u97f3, \u8f38\u5165", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "mandarin text to speech, pitch, prosodic, speech, synthesis, \u6587\u53e5\u7ffb\u8a9e\u97f3, \u5408\u6210", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "bilingual, machine translation, translation, \u6a5f\u5668\u7ffb\u8b6f", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "aspect, functional, lexical, lexical semantic, mandarin chinese, meaning, parsing, phrase, roles, semantic, semantics, syntactic, syntax, thematic, theory, verb, verbal, verbs", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| } |
| ], |
| "back_matter": [], |
| "bib_entries": { |
| "BIBREF0": { |
| "ref_id": "b0", |
| "title": "\u6a5f\u5668\u7ffb\u8b6f' 32 \u6a5f\u5668\u7ffb\u8b6f 'bilingual', 'machine translation', 'translation', '\u6a5f\u5668\u7ffb\u8b6f' 1991 \"Lexicon-Driven Transfer In English-Chinese Machine Translation\" 1992 \"A Modular and Statistical Approach to Machine Translation", |
| "authors": [ |
| { |
| "first": "Lee-Feng", |
| "middle": [], |
| "last": "Chien", |
| "suffix": "" |
| } |
| ], |
| "year": 1997, |
| "venue": "\u53ea\u6709\u8207\u53e2\u96c6 12 \u76f8\u95dc) 1995 \"THE NEW GENERATION BEHAVIORTRAN: DESIGN PHILOSOPHY AND SYSTEM ARCHITECTURE\" 1996 \"\u4ecb\u8a5e\u7ffb\u8b6f\u6cd5\u5247\u7684\u81ea\u52d5\u64f7\u53d6\" 2001 \"\u7d71\u8a08\u5f0f\u7247\u8a9e\u7ffb\u8b6f\u6a21\u578b", |
| "volume": "", |
| "issue": "", |
| "pages": "50--58", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "\u6a5f\u5668\u7ffb\u8b6f 'bilingual', 'machine translation', 'mt', 'transfer', '\u6a5f\u5668\u7ffb\u8b6f' 32 \u6a5f\u5668\u7ffb\u8b6f 'bilingual', 'machine translation', 'translation', '\u6a5f\u5668\u7ffb\u8b6f' 1991 \"Lexicon-Driven Transfer In English-Chinese Machine Translation\" 1992 \"A Modular and Statistical Approach to Machine Translation\" (\u53ea\u6709\u8207\u53e2\u96c6 12 \u76f8\u95dc) 1995 \"THE NEW GENERATION BEHAVIORTRAN: DESIGN PHILOSOPHY AND SYSTEM ARCHITECTURE\" 1996 \"\u4ecb\u8a5e\u7ffb\u8b6f\u6cd5\u5247\u7684\u81ea\u52d5\u64f7\u53d6\" 2001 \"\u7d71\u8a08\u5f0f\u7247\u8a9e\u7ffb\u8b6f\u6a21\u578b\" [Chien, 1997] Lee-Feng Chien, \"PAT-Tree-Based Keyword Extraction for Chinese Information Retrieval\", SIGIR'97, p50-58.", |
| "links": null |
| }, |
| "BIBREF1": { |
| "ref_id": "b1", |
| "title": "Recent Results on Domain-Specific Term Extraction From Online Chinese Text Resources", |
| "authors": [ |
| { |
| "first": "", |
| "middle": [], |
| "last": "Chien", |
| "suffix": "" |
| } |
| ], |
| "year": 1999, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "203--218", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Chien, et. al., 1999] Lee-Feng Chien, Chun-Liang Chen, Wen-Hsiang Lu, and Yuan-Lu Chang, \"Recent Results on Domain-Specific Term Extraction From Online Chinese Text Resources\", ROCLING XII, p203-218.", |
| "links": null |
| }, |
| "BIBREF2": { |
| "ref_id": "b2", |
| "title": "Introduction to the Special Issue on Computational Linguistics Using Large Corpora", |
| "authors": [ |
| { |
| "first": "K", |
| "middle": [ |
| "W" |
| ], |
| "last": "Church", |
| "suffix": "" |
| }, |
| { |
| "first": "R", |
| "middle": [ |
| "L" |
| ], |
| "last": "Mercer", |
| "suffix": "" |
| } |
| ], |
| "year": 1993, |
| "venue": "Computational Linguistics", |
| "volume": "19", |
| "issue": "1", |
| "pages": "1--24", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Church and Mercer, 1993] K. W. Church and R. L. Mercer, \"Introduction to the Special Issue on Computational Linguistics Using Large Corpora\", Computational Linguistics, 19(1), p1-24.", |
| "links": null |
| }, |
| "BIBREF3": { |
| "ref_id": "b3", |
| "title": "Material Mastery: Situating Digital Library Use in University Research Practices", |
| "authors": [ |
| { |
| "first": "L", |
| "middle": [ |
| "M" |
| ], |
| "last": "Covi", |
| "suffix": "" |
| } |
| ], |
| "year": null, |
| "venue": "", |
| "volume": "35", |
| "issue": "", |
| "pages": "293--316", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Covi, 1999] L. M. Covi, \"Material Mastery: Situating Digital Library Use in University Research Practices\", Information Processing and Management, 35, p293-316.", |
| "links": null |
| }, |
| "BIBREF4": { |
| "ref_id": "b4", |
| "title": "Indexing by Latent Semantic Analysis", |
| "authors": [ |
| { |
| "first": "", |
| "middle": [], |
| "last": "Deerwester", |
| "suffix": "" |
| } |
| ], |
| "year": 1990, |
| "venue": "Journal of the American Society for Information Science", |
| "volume": "41", |
| "issue": "6", |
| "pages": "391--407", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Deerwester, et. al., 1990] S. Deerwester, S. T. Dumais, G. W. Furnas, T. K. Landauer, and R. Harshman, \"Indexing by Latent Semantic Analysis\", Journal of the American Society for Information Science, 41(6), p391-407.", |
| "links": null |
| }, |
| "BIBREF5": { |
| "ref_id": "b5", |
| "title": "Psychological Relevance and Information Science", |
| "authors": [ |
| { |
| "first": "S", |
| "middle": [ |
| "P" |
| ], |
| "last": "Harter", |
| "suffix": "" |
| } |
| ], |
| "year": 1992, |
| "venue": "Journal of the American Society for Information Science", |
| "volume": "43", |
| "issue": "9", |
| "pages": "602--615", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Harter, 1992] S. P. Harter, \"Psychological Relevance and Information Science\", Journal of the American Society for Information Science, 43(9), p602-615.", |
| "links": null |
| }, |
| "BIBREF6": { |
| "ref_id": "b6", |
| "title": "An Investigation of Linguistic Features and Clustering Algorithms for Topical Document Clustering", |
| "authors": [ |
| { |
| "first": "[", |
| "middle": [], |
| "last": "Hatzivassiloglou", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Gravano", |
| "suffix": "" |
| }, |
| { |
| "first": ";", |
| "middle": [ |
| "V" |
| ], |
| "last": "Maganti", |
| "suffix": "" |
| }, |
| { |
| "first": "L", |
| "middle": [], |
| "last": "Hatzivassiloglou", |
| "suffix": "" |
| }, |
| { |
| "first": "A", |
| "middle": [], |
| "last": "Gravano", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Maganti", |
| "suffix": "" |
| } |
| ], |
| "year": 2000, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "224--231", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Hatzivassiloglou, Gravano and Maganti, 2000] V. Hatzivassiloglou, L. Gravano and A. Maganti, \"An Investigation of Linguistic Features and Clustering Algorithms for Topical Document Clustering\", SIGIR'2000, p224-231.", |
| "links": null |
| }, |
| "BIBREF7": { |
| "ref_id": "b7", |
| "title": "Towards a New Horizon in Information Science: Domain-Analysis", |
| "authors": [ |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Hj\u00f8rland", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [], |
| "last": "Albrechtsen", |
| "suffix": "" |
| } |
| ], |
| "year": 1995, |
| "venue": "", |
| "volume": "46", |
| "issue": "", |
| "pages": "400--425", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Hj\u00f8rland and Albrechtsen, 1995] B. Hj\u00f8rland and H. Albrechtsen, \"Towards a New Horizon in Information Science: Domain-Analysis\", Journal of the American Society for Information Science, 46(6), p400-425.", |
| "links": null |
| }, |
| "BIBREF8": { |
| "ref_id": "b8", |
| "title": "From Quantitative to Qualitative Studies: Developments in Chinese Computational and Corpus Linguistics", |
| "authors": [ |
| { |
| "first": "Chu-Ren", |
| "middle": [], |
| "last": "Huang", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Huang", |
| "suffix": "" |
| } |
| ], |
| "year": 2000, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "473--509", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Huang, 2000] Chu-Ren Huang, \"From Quantitative to Qualitative Studies: Developments in Chinese Computational and Corpus Linguistics\", \u6f22\u5b78\u7814\u7a76, \u7b2c\u5341\u516b\u5377\u7279\u520a, p473-509.", |
| "links": null |
| }, |
| "BIBREF9": { |
| "ref_id": "b9", |
| "title": "Document and Term Clustering", |
| "authors": [ |
| { |
| "first": "[", |
| "middle": [], |
| "last": "Kowalski", |
| "suffix": "" |
| }, |
| { |
| "first": "; G", |
| "middle": [ |
| "J" |
| ], |
| "last": "Maybury", |
| "suffix": "" |
| }, |
| { |
| "first": "M", |
| "middle": [ |
| "T" |
| ], |
| "last": "Kowalski", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Maybury", |
| "suffix": "" |
| } |
| ], |
| "year": 2000, |
| "venue": "Information Storage and Retrieval Systems: Theory and Implementation", |
| "volume": "2", |
| "issue": "", |
| "pages": "139--163", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Kowalski and Maybury, 2000] G. J. Kowalski and M. T. Maybury, \"Document and Term Clustering\", Information Storage and Retrieval Systems: Theory and Implementation, 2 nd ed., Chapter 6, p139-163.", |
| "links": null |
| }, |
| "BIBREF10": { |
| "ref_id": "b10", |
| "title": "Past and Future Goals of Computational Linguistics", |
| "authors": [ |
| { |
| "first": "W", |
| "middle": [], |
| "last": "Lenders", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Rocling Xiv", |
| "suffix": "" |
| } |
| ], |
| "year": null, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "213--236", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Lenders, 2001] W. Lenders, \"Past and Future Goals of Computational Linguistics\", ROCLING XIV, p213-236.", |
| "links": null |
| }, |
| "BIBREF11": { |
| "ref_id": "b11", |
| "title": "Information Epidemics and the Growth of Physics", |
| "authors": [ |
| { |
| "first": "A", |
| "middle": [ |
| "N" |
| ], |
| "last": "Tabah", |
| "suffix": "" |
| } |
| ], |
| "year": null, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[Tabah, 1996] A. N. Tabah, Information Epidemics and the Growth of Physics, Ph. D. Dissertation of McGill University, Canada.", |
| "links": null |
| }, |
| "BIBREF12": { |
| "ref_id": "b12", |
| "title": "Topic Detection and Tracking in English and Chinese", |
| "authors": [ |
| { |
| "first": ";", |
| "middle": [ |
| "C L" |
| ], |
| "last": "Wayne", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Wayne", |
| "suffix": "" |
| } |
| ], |
| "year": 2000, |
| "venue": "IRAL", |
| "volume": "5", |
| "issue": "", |
| "pages": "165--172", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Wayne, 2000] C. L. Wayne, \"Topic Detection and Tracking in English and Chinese\", IRAL 5, p165-172.", |
| "links": null |
| }, |
| "BIBREF13": { |
| "ref_id": "b13", |
| "title": "Models in Information Behaviour Research", |
| "authors": [ |
| { |
| "first": ";", |
| "middle": [ |
| "T D" |
| ], |
| "last": "Wilson", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Wilson", |
| "suffix": "" |
| } |
| ], |
| "year": 1999, |
| "venue": "Journal of Documentation", |
| "volume": "55", |
| "issue": "3", |
| "pages": "249--270", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Wilson, 1999] T. D. Wilson, \"Models in Information Behaviour Research\", Journal of Documentation, 55(3), p249-270.", |
| "links": null |
| }, |
| "BIBREF14": { |
| "ref_id": "b14", |
| "title": "A Study on Retrospective and On-Line Event Detection", |
| "authors": [ |
| { |
| "first": "Pierce", |
| "middle": [], |
| "last": "Yang", |
| "suffix": "" |
| }, |
| { |
| "first": ";", |
| "middle": [ |
| "Y" |
| ], |
| "last": "Carbonell", |
| "suffix": "" |
| }, |
| { |
| "first": "T", |
| "middle": [], |
| "last": "Yang", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Pierce", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Carbonell", |
| "suffix": "" |
| } |
| ], |
| "year": 1998, |
| "venue": "SIGIR'98", |
| "volume": "", |
| "issue": "", |
| "pages": "28--36", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Yang, Pierce and Carbonell, 1998] Y. Yang, T. Pierce, and J. Carbonell, \"A Study on Retrospective and On-Line Event Detection\", SIGIR'98, p28-36.", |
| "links": null |
| }, |
| "BIBREF15": { |
| "ref_id": "b15", |
| "title": "\u96fb\u8166\u5728\u8a9e\u8a00\u5b78\u88e1\u7684\u904b\u7528", |
| "authors": [ |
| { |
| "first": "", |
| "middle": [], |
| "last": "Rocling I", |
| "suffix": "" |
| } |
| ], |
| "year": 1988, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "257--287", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "[\u738b\u58eb\u5143, 1988] \"\u96fb\u8166\u5728\u8a9e\u8a00\u5b78\u88e1\u7684\u904b\u7528\", ROCLING I, p257-287.", |
| "links": null |
| } |
| }, |
| "ref_entries": { |
| "TABREF6": { |
| "html": null, |
| "type_str": "table", |
| "content": "<table><tr><td>\u51fa\u73fe\u7e3d\u983b\u6b21\u6700\u9ad8\u7684\u524d 50 \u500b\u8a5e\u8a9e\u53ca\u51fa\u73fe\u7684\u7e3d\u983b\u6b21\u5217\u8868\u65bc\u8868\u4e00\u3002 \u8868\u4e00 \u95dc\u9375\u8a5e\u8a9e\u62bd\u53d6\u6240\u5f97\u5230\u7684\u524d 50 \u500b\u51fa\u73fe\u7e3d\u983b\u6b21\u6700\u9ad8\u7684\u8a5e\u8a9e\u53ca\u7e3d\u983b\u6b21 \u4f75\u5f8c\uff0c\u6240\u5f97\u5230\u4e09\u500b\u8a5e\u8a9e\u4ee5\u4e0a\u7684\u53e2\u96c6\u7684\u6578\u76ee\uff0c\u5982\u8868\u4e8c\u6240\u793a\u3002 \u8868\u4e8c \u4e0d\u540c\u76f8\u95dc\u7a0b\u5ea6\u4f30\u7b97\u65b9\u6cd5\u9032\u884c\u7814\u7a76\u4e3b\u984c\u53e2\u96c6\u6240\u5f97\u5230\u7684\u53e2\u96c6\u6578\u76ee \u8a9e\u3002\u56e0\u6b64\uff0c\u672c\u8ad6\u6587\u6240\u63d0\u51fa\u4f86\u7684\u7814\u7a76\u4e3b\u984c\u62bd\u53d6\u65b9\u6cd5\u7684\u53ef\u884c\u6027\u4fbf\u53ef\u4ee5\u5f97\u5230\u521d\u6b65\u9a57\u8b49\u3002 \u8868\u4e09 \u8207\u8a9e\u8a00\u7684\u8a08\u7b97\u6a21\u5f0f\u76f8\u95dc\u7684\u8a5e\u8a9e\u53e2\u96c6\u53ca\u76f8\u95dc\u8ad6\u6587 \u6b64\u5916\u5f9e\u7814\u7a76\u7d50\u679c\u4e2d\uff0c\u6211\u5011\u4e5f\u767c\u73fe\u8a08\u7b97\u8a9e\u8a00\u5b78\u7814\u7a76\u8207\u5be6\u52d9\u61c9\u7528\u6709\u5bc6\u5207\u7684\u95dc\u4fc2\uff0c \u8868\u56db\u5230\u8868\u516d\u5206\u5225\u5217\u51fa\u8207\u6a5f\u5668\u7ffb\u8b6f\u3001\u8a9e\u97f3\u8655\u7406\u548c\u8cc7\u8a0a\u6aa2\u7d22\u76f8\u95dc\u7684\u96c6\u5408\u3002\u5f9e\u8868\u56db\u7684\u7d50 dictation, 1993 \"\u570b\u8a9e\u8a9e\u97f3\u8fa8\u8a8d\u4e2d\u8a5e\u7fa4\u96d9\u9023\u8a9e\u8a00\u6a21\u578b\u7684\u89e3\u78bc\u65b9\u6cd5\" csmart, databases, 1995 \"\u9069\u5408\u5927\u91cf\u4e2d\u6587\u6587\u4ef6\u5168\u6587\u6aa2\u7d22\u7684\u7d22\u5f15\u53ca\u8cc7\u6599\u58d3\u7e2e\u6280\u8853\" \u6a21\u578b\u5247\u662f\u570b\u5167\u8a08\u7b97\u8a9e\u8a00\u5b78\u5bb6\u6240\u95dc\u5fc3\u7684\u4e3b\u984c\u3002 \u5728\u5f8c\u7e8c\u7684\u7814\u7a76\u4e0a\uff0c\u9664\u4e86\u9032\u4e00\u6b65\u6539\u5584\u76ee\u524d\u6240\u63d0\u51fa\u4f86\u7684\u65b9\u6cd5\uff0c\u4e26\u4e14\u6df1\u5165\u63a2\u8a0e\u5404\u7814 \u5426\u51fa\u73fe\u5728\u53e2\u96c6\u03a7\u4e4b\u4e2d\uff0c\u63db\u8a00\u4e4b\uff0c\u5982\u679c\u7b2c i \u884c\u5411\u91cf\uff0c\u5927\u5c0f\u70ba 1\u00d7d\uff0c\u6bcf\u4e00\u500b\u5143\u7d20\u6240\u4ee3\u8868\u7684\u503c\u70ba\u8a5e\u8a9e\u53e2\u96c6\u03a7\u8207\u6240\u5c0d\u61c9\u7684\u8ad6\u6587\u4e4b\u9593\u7684 \u76f8\u95dc\u7a0b\u5ea6\u4f30\u7b97\u503c\u3002\u6700\u5f8c\u4f9d\u64da\u9019\u500b\u7d50\u679c\uff0c\u5c07\u76f8\u95dc\u7a0b\u5ea6\u5927\u7684\u8ad6\u6587\u8cc7\u6599\u53d6\u51fa\uff0c\u4f5c\u70ba\u7814\u7a76 \u4e3b\u984c\u76f8\u95dc\u7684\u8ad6\u6587\u8cc7\u6599\u4f86\u9032\u884c\u5206\u6790\u3002 \u4e94\u3001\u570b\u5167\u8a08\u7b97\u8a9e\u8a00\u5b78\u7684\u7814\u7a76\u4e3b\u984c\u5206\u6790\u7684\u5be6\u9a57\u7d50\u679c \u8a08\u7b97\u8a9e\u8a00\u5b78\u7814\u8a0e\u6703 ROCLING \u662f\u570b\u5167\u7684\u8a08\u7b97\u8a9e\u8a00\u5b78\u9818\u57df\u76f8\u7576\u91cd\u8981\u7684\u5b78\u8853\u6d3b \u52d5\u3002\u56e0\u6b64\uff0cROCLING \u7684\u7814\u8a0e\u6703\u8ad6\u6587\u96c6\u4e2d\u8ad6\u6587\u8cc7\u6599\uff0c\u53ef\u4ee5\u8aaa\u662f\u6b77\u5e74\u4f86\u570b\u5167\u8a08\u7b97\u8a9e\u8a00 \u5b78\u9818\u57df\u5b78\u8005\u7684\u5fc3\u8840\u7d50\u6676\uff0c\u6240\u860a\u542b\u7684\u7814\u7a76\u4e3b\u984c\u4e5f\u662f\u4ed6\u5011\u6240\u95dc\u5fc3\u7684\u7814\u7a76\u4e3b\u984c\u3002\u56e0\u6b64\uff0c \u672c\u8ad6\u6587\u5c07\u4ee5 ROCLING \u7814\u8a0e\u6703\u7684\u8ad6\u6587\u8cc7\u6599\u505a\u70ba\u5206\u6790\u570b\u5167\u8a08\u7b97\u8a9e\u8a00\u5b78\u7814\u7a76\u4e3b\u984c\u7684\u7d20 \u6750\u3002 \u5206\u6790\u8cc7\u6599\u70ba\u5f9e\u7b2c\u4e00\u5c46(1988)\u5230\u7b2c\u5341\u56db\u5c46(2001)\u7684 ROCLING \u7814\u8a0e\u6703\u8ad6\u6587\uff0c\u5171 235 \u7bc7\u3002\u9032\u884c\u8a5e\u8a9e\u62bd\u53d6\u6642\uff0c\u9996\u5148\u62bd\u53d6\u91cd\u8981\u7684\u591a\u5b57\u8a5e\u53ca\u8a5e\u7d44\uff0c\u6240\u8a2d\u5b9a\u7684\u5b57\u4e32\u51fa\u73fe\u7e3d\u983b\u6b21 \u6b21\u5e8f \u8a5e\u540d \u51fa\u73fe\u7e3d\u983b\u6b21 \u6b21\u5e8f \u8a5e\u540d \u51fa\u73fe\u7e3d\u983b\u6b21 1 parsing 209 26 parser 80 2 speech 184 27 probabilistic 78 3 \u7cfb\u7d71 175 28 \u52d5\u8a5e 78 4 sentences 141 29 \u8a9e\u97f3 78 5 lexical 138 30 knowledge 74 6 mandarin 134 31 \u8a9e\u6cd5 74 7 speech recognition 132 32 chinese text 73 8 \u65b9\u6cd5 131 33 \u8a9e\u8a00 73 9 semantic 130 34 semantics 72 10 corpus 129 35 corpora 71 11 syntactic 107 36 used 71 12 recognition 106 37 \u570b\u8a9e 71 13 data 105 38 discourse 70 14 \u5206\u6790 104 39 \u8655\u7406 70 15 learning 102 40 dictionary 68 16 mandarin chinese 97 41 problem 65 17 sentence 97 42 \u5206\u985e 65 18 machine translation 92 43 corpus based 64 19 words 92 44 design 62 20 theory 87 45 information retrieval 62 Original feature vectors SVD k=120 SVD k=60 SVD k=30 cliques \u53e2\u96c6 65 78 85 74 \u53e2\u96c6\u5408\u4f75 27 34 34 32 \u5f9e\u8868\u4e8c\u4e2d\uff0c\u53ef\u4ee5\u89c0\u5bdf\u5230\u7d93\u904e SVD \u8655\u7406\u7684 cliques \u53e2\u96c6\u6578\u76ee\u8f03\u539f\u5148\u7684\u7279\u5fb5\u5411\u91cf\u4f86 \u5f97\u591a\uff0c\u986f\u7136 LSI \u6280\u8853\u6709\u52a9\u65bc\u6355\u6349\u8a5e\u8a9e\u4e0d\u5171\u73fe\u537b\u76f8\u95dc\u7684\u96b1\u542b\u8a9e\u7fa9\u7d50\u69cb\uff0c\u7522\u751f\u8f03\u591a cliques \u53e2\u96c6\u3002\u56e0\u6b64\uff0c\u6211\u5011\u4ee5\u7d93 SVD \u8655\u7406 k \u503c\u70ba 60 \u7684\u7279\u5fb5\u5411\u91cf\u9032\u884c\u8a5e\u8a9e\u76f8\u95dc\u7a0b\u5ea6 \u4f30\u7b97\uff0c\u5c07\u6240\u5f97\u5230\u7684 34 \u500b\u8a5e\u8a9e\u53e2\u96c6\u4f5c\u70ba\u9032\u4e00\u6b65\u7684\u5206\u6790\u7684\u5c0d\u8c61\uff0c\u9019 34 \u500b\u8a5e\u8a9e\u53e2\u96c6\u5217 \u8868\u65bc\u9644\u9304\u4e00\u3002 \u5f9e\u8a5e\u8a9e\u53e2\u96c6\u7684\u7d50\u679c\u6211\u5011\u53ef\u4ee5\u770b\u5230\u5e7e\u500b\u73fe\u8c61\u3002\u7b2c\u4e00\u3001\u82e5\u5e72\u53e2\u96c6\u540c\u6642\u5177\u6709\u4e2d\u6587\u8a5e n gram \u806f\u6027\" \u4e2d\u6240\u5f97\u5230\u7684\u7d50\u679c\u53ef\u4ee5\u5206\u6790\u6210\u8a9e\u8a00\u6a21\u578b\u3001\u8072\u5b78\u8fa8\u8a8d\u4ee5\u53ca\u8a9e\u97f3\u5408\u6210\u4e09\u500b\u7814\u7a76\u4e3b\u984c(\u8868 \u5728\u8a08\u7b97\u8a9e\u8a00\u5b78\u9818\u57df\u4e2d\uff0c\u8cc7\u8a0a\u6aa2\u7d22\u6bd4\u8d77\u5176\u4ed6\u7814\u7a76\u53ef\u8aaa\u662f\u4e00\u500b\u8f03\u65b0\u7684\u4e3b\u984c\uff0c\u7136\u800c \u6b64\u9593\u5171\u73fe\u95dc\u4fc2\u9032\u884c\u53e2\u96c6\uff0c\u4ee5\u53e2\u96c6\u6240\u5f97\u5230\u7684\u8a5e\u8a9e\u96c6\u5408\u8868\u793a\u9818\u57df\u4e2d\u91cd\u8981\u7684\u7814\u7a76\u4e3b\u984c\u3002 \u542b\u4e86'machine translation'\u3001'mt'\u3001'\u6a5f\u5668\u7ffb\u8b6f'\u7b49\u8a5e\u8a9e\uff1b\u6216\u662f\u53c8\u5982\u53e2\u96c6 18 \u5305\u542b\u4e86 'word \u5efa\u7acb language modeling, language models, Language Model Using Kullback-Leibler Distance Criterion\" 2001 \"\u4f7f\u7528\u95dc\u806f\u6cd5\u5247\u70ba\u4e3b\u4e4b\u8a9e\u8a00\u6a21\u578b\u65bc\u64f7\u53d6\u9577\u8ddd\u96e2\u4e2d\u6587\u6587\u5b57\u95dc \u4f86\u8a9e\u97f3\u8655\u7406\u5df2\u7d93\u6210\u70ba\u8a08\u7b97\u8a9e\u8a00\u5b78\u76f8\u7576\u91cd\u8996\u7684\u7814\u7a76\u4e3b\u984c\u3002\u5f9e ROCLING \u7684\u8ad6\u6587\u8cc7\u6599 synthesis, \u6587\u53e5\u7ffb\u8a9e\u97f3, \u5408\u6210 2001 \"Pitch Marking Based on an Adaptable Filter and a Peak-Valley Estimation Method\", (\u53ea\u6709\u8207\u53e2\u96c6 31 \u76f8\u95dc) \u8853\uff0c\u5f9e\u5b78\u8853\u9818\u57df\u4e2d\u767c\u8868\u7684\u8ad6\u6587\u8cc7\u6599\u4e2d\u62bd\u53d6\u91cd\u8981\u7684\u95dc\u9375\u8a5e\u8a9e\uff0c\u4e26\u5c07\u9019\u4e9b\u8a5e\u8a9e\u4f9d\u64da\u5f7c \u8a9e\u8207\u82f1\u6587\u8a5e\u8a9e\uff0c\u751a\u81f3\u5305\u542b\u7e2e\u5beb\u8207\u76f8\u540c\u6982\u5ff5\u4f46\u4e0d\u540c\u8a5e\u540d\u7684\u8a5e\u8a9e\uff0c\u6bd4\u65b9\u8aaa\uff0c\u53e2\u96c6 12 \u5305 \u8a9e\u6cd5\u6a21\u5f0f \u8207\u5256\u6790 \u5206\u6790, \u8868\u9054, \u5256\u6790, \u683c\u4f4d, \u8a0a\u606f, \u52d5\u8a5e, \u7d50\u69cb, \u8a5e\u985e, \u6f22\u8a9e, \u8a9e\u6cd5, \u8a9e\u6cd5\u6a21\u5f0f, \u8a9e\u610f, \u6a21\u5f0f, \u95dc\u4fc2 \u679c\uff0c\u8aaa\u660e\u6a5f\u5668\u7ffb\u8b6f\u662f\u8a08\u7b97\u8a9e\u8a00\u5b78\u6700\u65e9\u7684\u61c9\u7528\u554f\u984c\u4e4b\u4e00[Lenders, 2001]\uff0c\u800c\u5176\u767c\u5c55\u5f9e \u8a9e\u8a00\u6a21\u578b large vocabulary, 1994 \"\u570b\u8a9e\u8a9e\u97f3\u8fa8\u8a8d\u4e2d\u8a5e\u7fa4\u8a9e\u8a00\u6a21\u578b\u4e4b\u5206\u7fa4\u65b9\u6cd5\u8207\u61c9\u7528\" \u8cc7\u8a0a\u6aa2\u7d22 document, indexing, 1996 \"\u5c0b\u6613(Csmart-II):\u667a\u6167\u578b\u7db2\u8def\u4e2d\u6587\u8cc7\u8a0a\u6aa2\u7d22\u7cfb\u7d71\" \u7a76\u4e3b\u984c\u7684\u8d77\u6e90\u3001\u767c\u5c55\u8207\u6f14\u8b8a\u4e4b\u5916\uff0c\u6211\u5011\u5c07\u63a2\u7d22\u5404\u500b\u7814\u7a76\u4e3b\u984c\u4e4b\u9593\u7684\u76f8\u95dc\u6027\uff0c\u4e26\u5617 \u8a9e\u8a00\u6a21\u578b, \u8a9e\u97f3\u8fa8\u8a8d 1995 \"\u61c9\u7528\u65bc'\u97f3\u4e2d\u4ed9'\u570b\u8a9e\u807d\u5beb\u6a5f\u4e4b\u77ed\u8a9e\u898f\u5247\u5206\u6790\u8207\u5efa\u7acb\" information retrieval, 1997 \"An Assessment on Character-based Chinese News Filtering 1989 \"\u8a0a\u606f\u70ba\u672c\u7684\u683c\u4f4d\u8a9e\u6cd5--\u4e00\u500b\u9069\u7528\u65bc\u8868\u9054\u4e2d\u6587\u7684\u8a9e\u6cd5\u6a21\u5f0f\" 1991 \"\u9023\u63a5\u8a5e\u7684\u8a9e\u6cd5\u8868\u9054\u6a21\u5f0f-\u4ee5\u4e2d\u6587\u8a0a\u606f\u683c\u4f4d\u8a9e\u6cd5(ICG)\u70ba\u672c \u7684\u8868\u9054\u5f62\u5f0f\" 1992 \"\u6f22\u8a9e\u7684\u52d5\u8a5e\u540d\u7269\u5316\u521d\u63a2--\u6f22\u8a9e\u4e2d\u5e36\u8ad6\u5143\u7684\u540d\u7269\u5316\u6d3e\u751f\u540d \u8a5e\" 18 \u65b7\u8a5e chinese text, chinese word segmentation, segmentation, unknown word, word identification, word segmentation, words, \u65b7\u8a5e 1994 \"Chinese-Word Segmentation Based on Maximal-Matching and Bigram Techniques\" 1995 \"A Unifying Approach to Segmentation of Chinese and Its Application to Text Retrieval\" 1997 \"Unknown Word Detection for Chinese by a Corpus-based Learning Method\" 1997 \"Chinese Word Segmentation and Part-of-Speech Tagging in One Step\" 1997 \"A Simple Heuristic Approach for Word Segmentation\" 22 \u7d71\u8a08\u5f0f\u8a9e \u8a00\u6a21\u578b\u7684 bigram, class based, clustering, entropy, language model, 1994 \"An Estimation of the Entropy of Chinese -A New Approach to Constructing Class-based n-gram Models\" 1997 \"Truncation on Combined Word-Based and Class-Based \u898f\u5247\u5f0f\u7684\u81ea\u52d5\u7ffb\u8b6f\u5230\u7d71\u8a08\u5f0f\uff0c\u8fd1\u671f\u7684\u61c9\u7528\u5247\u662f\u5728\u8de8\u8a9e\u8a00\u6aa2\u7d22\u90e8\u5206\u3002 \u8868\u56db \u8207\u6a5f\u5668\u7ffb\u8b6f\u76f8\u95dc\u7684\u8a5e\u8a9e\u53e2\u96c6\u53ca\u76f8\u95dc\u8ad6\u6587 \u5728\u904e\u53bb\u8a08\u7b97\u8a9e\u8a00\u5b78\u6240\u8655\u7406\u7684\u5c0d\u8c61\u591a\u70ba\u66f8\u5beb\u8a9e\u8a00(orthographic languages)\uff0c\u8fd1\u5e74 17 \u8a9e\u8a00\u6a21\u578b \u570b\u8a9e, \u8a9e\u8a00\u6a21\u578b, \u8a9e\u97f3\u8fa8\u8a8d, \u8fa8\u8a8d 1996 \"\u570b\u8a9e\u8a9e\u97f3\u8fa8\u8a8d\u4e2d\u591a\u9818\u57df\u8a9e\u8a00\u6a21\u578b\u4e4b\u8a13\u7df4\u3001\u5075\u6e2c\u8207\u8abf\u9069\" 1999 \"\u570b\u8a9e\u96fb\u8a71\u8a9e\u97f3\u8fa8\u8a8d\u4e4b\u5f37\u5065\u6027\u7279\u5fb5\u53c3\u6578\u53ca\u5176\u8abf\u6574\u65b9\u6cd5\" (\u53ea\u6709\u8207\u53e2\u96c6 17 \u76f8\u95dc) 7 \u8072\u5b78\u8fa8\u8a8d hidden markov, maximum, robust speech recognition, speech recognition 1998 \"Speaker-Independent Continuous Mandarin Speech Recognition Under Telephone Environments\" 1999 \"\u570b\u8a9e\u96fb\u8a71\u8a9e\u97f3\u8fa8\u8a8d\u4e4b\u5f37\u5065\u6027\u7279\u5fb5\u53c3\u6578\u53ca\u5176\u8abf\u6574\u65b9\u6cd5\" 2000 \"\u5177\u6709\u7d2f\u9032\u5b78\u7fd2\u80fd\u529b\u4e4b\u8c9d\u6c0f\u9810\u6e2c\u6cd5\u5247\u5728\u6c7d\u8eca\u8a9e\u97f3\u8fa8\u8b58\u4e4b\u61c9 \u7528\" 2000 \"\u7d9c\u5408\u9ea5\u514b\u98a8\u9663\u5217\u53ca\u6a21\u578b\u8abf\u6574\u6280\u8853\u4e4b\u9060\u8ddd\u96e2\u8a9e\u97f3\u8fa8\u8b58\u7cfb\u7d71\" 30 \u8a9e\u97f3\u5408\u6210 speech, synthesis, \u6587\u53e5\u7ffb\u8a9e\u97f3, \u5408\u6210, \u7cfb\u7d71, \u97f3\u7bc0, \u570b\u8a9e, \u9023\u97f3, \u8a9e\u97f3, \u8f38\u5165 31 \u8a9e\u97f3\u5408\u6210 mandarin text to speech, pitch, prosodic, speech, 1995 \"\u4ee5 CELP \u70ba\u57fa\u790e\u4e4b\u6587\u53e5\u7ffb\u8a9e\u97f3\u4e2d\u97fb\u5f8b\u8a0a\u606f\u4e4b\u7522\u751f\u8207\u8abf\u6574\" 1996 \"\u6642\u9593\u6bd4\u4f8b\u57fa\u9031\u6ce2\u5f62\u5167\u5dee--\u4e00\u500b\u570b\u8a9e\u97f3\u7bc0\u4fe1\u865f\u5408\u6210\u4e4b\u65b0\u65b9 \u6cd5\" 1996 \"\u4e2d\u82f1\u6587\u6587\u53e5\u7ffb\u8a9e\u97f3\u7cfb\u7d71\u4e2d\u9023\u97f3\u8655\u7406\u4e4b\u7814\u7a76\" 1999 \"\u53f0\u8a9e\u591a\u8072\u8abf\u97f3\u7bc0\u5408\u6210\u55ae\u5143\u8cc7\u6599\u5eab\u66a8\u6587\u5b57\u8f49\u8a9e\u97f3\u96db\u5f62\u7cfb\u7d71 \u4e4b\u767c\u5c55\" (\u53ea\u6709\u8207\u53e2\u96c6 30 \u76f8\u95dc) 1999 \"\u570b\u8a9e\u6587\u53e5\u7ffb\u53f0\u8a9e\u8a9e\u97f3\u7cfb\u7d71\u4e4b\u7814\u7a76\" (\u53ea\u6709\u8207\u53e2\u96c6 30 \u76f8\u95dc) retrieval, text retrieval, \u6aa2\u7d22 Using Latent Semantic Indexing\" 1999 \"A New Syllable-Based Approach for Retrieving Mandarin Spoken Documents Using Short Speech Queries\" 9 \u6587\u4ef6\u5206\u985e document, hierarchical, text categorization, \u5206\u985e, \u6587\u4ef6, \u6587\u4ef6\u5206\u985e, \u7279\u5fb5 28 \u6587\u4ef6\u5206\u985e document, text categorization, \u5206\u985e, \u6587\u4ef6\u5206\u985e, \u6587\u4ef6\u81ea\u52d5, \u95dc\u9375\u8a5e 1993 \"\u4e2d\u6587\u6587\u4ef6\u81ea\u52d5\u5206\u985e\u4e4b\u7814\u7a76\" 1999 \"\u968e\u5c64\u5f0f\u6587\u4ef6\u81ea\u52d5\u5206\u985e\u4e4b\u7279\u5fb5\u9078\u53d6\u7814\u7a76\" 2001 \"\u57fa\u65bc\u968e\u5c64\u5f0f\u795e\u7d93\u7db2\u8def\u4e4b\u81ea\u52d5\u6587\u4ef6\u5206\u985e\u65b9\u6cd5\" 2001 \"\u9069\u61c9\u6027\u6587\u4ef6\u5206\u985e\u7cfb\u7d71\" \u516d\u3001\u7d50\u8ad6 \u672c\u8ad6\u6587\u91dd\u5c0d\u7814\u7a76\u4e3b\u984c\u5206\u6790\u7684\u554f\u984c\uff0c\u63d0\u51fa\u4e00\u7cfb\u5217\u4ee5\u81ea\u7136\u8a9e\u8a00\u8655\u7406\u70ba\u57fa\u790e\u7684\u6280 \u8a66\u5c07\u7d50\u679c\u4ee5\u5716\u5f62\u5316\u7684\u65b9\u5f0f\u52a0\u4ee5\u5448\u73fe\u3002\u53e6\u5916\uff0c\u5c0d\u65bc\u4e0d\u540c\u5b78\u8853\u9818\u57df\u9593\u7684\u76f8\u95dc\u7814\u7a76\u4e3b\u984c \u7684\u767c\u6398\u548c\u5206\u6790\uff0c\u6bd4\u65b9\u8aaa\u8cc7\u8a0a\u6aa2\u7d22\u540c\u6a23\u662f\u5716\u66f8\u8cc7\u8a0a\u5b78\u6240\u95dc\u5fc3\u7684\u7814\u7a76\u4e3b\u984c\uff0c\u5982\u4f55\u5229\u7528 \u81ea\u7136\u8a9e\u8a00\u8655\u7406\u6280\u8853\u4f86\u5206\u6790\u5169\u500b\u9818\u57df\u9593\u7684\u5171\u901a\u8207\u76f8\u7570\uff0c\u662f\u4e00\u9805\u503c\u5f97\u63a2\u8a0e\u7684\u7814\u7a76\u3002 \u81f4\u8b1d \u672c \u7814 \u7a76 \u53d7 \u570b \u79d1 \u6703 \u300c \u570b \u5167 \u8a08 \u7b97 \u8a9e \u8a00 \u5b78 \u5b78 \u8853 \u8cc7 \u8a0a \u4ea4 \u6d41 \u4e4b \u7814 \u7a76 (I) \u300d ( \u7de8 \u865f NSC91-2413-H-128-004-)\u8a08\u756b\u6848\u88dc\u52a9\u3002\u53e6\u5916\uff0c\u4e5f\u611f\u8b1d\u4e09\u4f4d\u5be9\u67e5\u59d4\u54e1\u6240\u63d0\u4f9b\u7684\u610f\u898b\u3002 \u53c3\u8003\u6587\u737b classifiers, decision, non, symbols 11 \u5206\u6790, \u7cfb\u7d71, \u8655\u7406, \u8a9e\u8a00 12 bilingual, machine translation, mt, transfer, \u6a5f\u5668\u7ffb\u8b6f 13 dictation, large vocabulary, \u8a9e\u8a00\u6a21\u578b, \u8a9e\u97f3\u8fa8\u8a8d 14 adaptation, maximum, robust speech recognition, \u8a9e\u97f3\u8fa8\u8b58 15 attachment, pp, preference, score 16 \u7cfb\u7d71, \u8a2d\u8a08, \u8f38\u5165, \u9375\u76e4 17 \u570b\u8a9e, \u8a9e\u8a00\u6a21\u578b, \u8a9e\u97f3\u8fa8\u8a8d, \u8fa8\u8a8d 18 chinese text, chinese word segmentation, segmentation, unknown word, word identification, word segmentation, words, \u65b7\u8a5e 19 attention, conversation, discourse, elicitation, interaction 20 continuous, hidden markov, maximum, speech recognition 21 \u7d71\u8a08, \u8a5e\u5f59, \u8a9e\u8a00, \u8a9e\u6599 22 bigram, class based, clustering, entropy, language model, language modeling, language models, n gram [Bishop, 1999] 10 23 \u5206\u6790,</td></tr><tr><td>21 rules 22 models identification'\u3001'word segmentation'\u3001'\u65b7\u8a5e'\u7b49\u8a5e\u8a9e\u3002\u53ef\u898b\u5c07\u53c3\u8003\u6587\u737b\u7684\u984c\u540d\u52a0\u5165\u8ad6 84 46 syntax 61 83 47 generation 60 \u7531\u65bc\u7bc7\u5e45\u7684\u9650\u5236\uff0c\u672c\u8ad6\u6587\u7121\u6cd5\u5c0d\u6240\u6709\u62bd\u53d6\u51fa\u4f86\u8a5e\u8a9e\u53e2\u96c6\u4e00\u4e00\u9032\u884c\u8a73\u76e1\u7684\u5831 \u4e94)\u3002\u570b\u5167\u8a08\u7b97\u8a9e\u8a00\u5b78\u8f03\u65e9\u9032\u884c\u7814\u7a76\u7684\u4e3b\u984c\u662f\u8a9e\u8a00\u6a21\u578b\u548c\u8a9e\u97f3\u5408\u6210\uff0c\u8fd1\u5e74\u5728\u8072\u5b78\u8fa8 \u5728\u672c\u8ad6\u6587\u4e2d\uff0c\u6211\u5011\u5c07\u6240\u63d0\u51fa\u7684\u65b9\u6cd5\u61c9\u7528\u5230 ROCLING \u7814\u8a0e\u6703\u7684\u8ad6\u6587\u8cc7\u6599\u4e0a\uff0c\u62bd\u53d6 \u7531\u65bc\u7db2\u969b\u7db2\u8def\u8207\u96fb\u5b50\u6587\u4ef6\u7684\u767c\u5c55\u4f7f\u5f97\u9019\u9805\u61c9\u7528\u6210\u70ba\u76f8\u7576\u5177\u6709\u6f5b\u529b\u7684\u7814\u7a76\u4e3b\u984c\u3002\u6211 \u7684\u95be\u503c\uff0c\u8f03\u77ed\u7684\u5b57\u4e32(2 \u6216 3 \u5b57)\u8a2d\u5b9a\u70ba 15 \u6b21\uff0c\u8f03\u9577\u7684\u5b57\u4e32(4~5 \u5b57)\u5247\u8a2d\u5b9a\u70ba 10 \u6b21\uff0c 23 phrase 83 48 \u8a9e\u6599\u5eab 60 24 \u6f22\u8a9e 82 49 \u61c9\u7528 60 \u6587\u8cc7\u6599\uff0c\u53ef\u4ee5\u7372\u5f97\u4e2d\u6587\u548c\u82f1\u6587\u5169\u7a2e\u8a9e\u8a00\u7684\u8a5e\u5f59\u8a0a\u606f\uff0c\u800c\u4e14\u5229\u7528\u8a5e\u8a9e\u7684\u5171\u73fe\u95dc\u4fc2\u53ef \u544a\uff0c\u4ee5\u4e0b\u91dd\u5c0d\u5e7e\u500b\u4e3b\u984c\u8f03\u660e\u78ba\u7684\u8a5e\u8a9e\u53e2\u96c6\u9032\u884c\u8aaa\u660e\u3002\u8868\u4e09\u662f\u8207\u8a9e\u8a00\u7684\u8a08\u7b97\u6a21\u5f0f\u76f8 \u8a8d\u7814\u7a76\u4e0a\uff0c\u4e5f\u6709\u8a31\u591a\u7814\u7a76\u4eba\u54e1\u9032\u5165\u9019\u500b\u9818\u57df\u767c\u8868\u76f8\u95dc\u8ad6\u6587\u3002\u5728\u8868\u4e94\uff0c\u53e6\u5916\u9084\u53ef\u5c07 \u8a08\u7b97\u8a9e\u8a00\u5b78\u9818\u57df\u7684\u91cd\u8981\u7814\u7a76\u4e3b\u984c\uff0c\u7d50\u679c\u986f\u793a\u9019\u500b\u65b9\u6cd5\u53ef\u4ee5\u540c\u6642\u62bd\u53d6\u51fa\u4e2d\u6587\u548c\u82f1\u6587 \u5011\u53ef\u4ee5\u5f9e\u8868\u516d\u4e2d\u767c\u73fe\u570b\u5167\u8a08\u7b97\u8a9e\u8a00\u5b78\u5728\u9019\u65b9\u9762\u7684\u91cd\u8981\u7814\u7a76\u5305\u62ec\u8cc7\u8a0a\u6aa2\u7d22\u548c\u6587\u4ef6\u5206 \u5b57\u4e32\u5c0d\u51fa\u73fe\u8ad6\u6587\u8cc7\u6599\u7684\u91cd\u8981\u7a0b\u5ea6 R S (\u5e73\u5747\u983b\u6b21\u548c\u6a19\u6e96\u5dee\u7684\u7e3d\u548c)\u8a2d\u70ba 2.5\uff0c\u524d\u5f8c\u63a5\u5b57 25 classification 80 50 character 59 \u4ee5\u5c07\u76f8\u95dc\u7684\u8a5e\u8a9e\u53e2\u805a\u8d77\u4f86\u3002\u7b2c\u4e8c\u3001\u5927\u90e8\u5206\u7684\u8a5e\u8a9e\u53e2\u805a\u90fd\u53ef\u4ee5\u660e\u986f\u5730\u7528\u4f86\u4ee3\u8868\u4e00\u500b \u95dc\u7684\u8a5e\u8a9e\u53e2\u96c6\u53ca\u76f8\u95dc\u8ad6\u6587\u7684\u5217\u8868\uff0c\u8ad6\u6587\u524d\u7684\u6578\u503c\u662f\u8ad6\u6587\u5728 ROCLING \u7814\u8a0e\u6703\u4e2d\u767c \u8a9e\u97f3\u5408\u6210\u7814\u7a76\u5206\u6210\u7cfb\u7d71\u88fd\u4f5c(\u53e2\u96c6 30)\u8207\u8072\u5b78\u8a0a\u606f\u7814\u7a76(\u53e2\u96c6 31)\u5169\u500b\u90e8\u5206\u3002 \u7684\u95dc\u9375\u8a5e\u8a9e\uff0c\u6240\u5f97\u5230\u7684\u8a5e\u8a9e\u53e2\u96c6\u7d50\u679c\u4e5f\u53ef\u4ee5\u8868\u793a\u9818\u57df\u4e2d\u91cd\u8981\u7684\u7814\u7a76\u4e3b\u984c\u3002\u9019\u6a23\u7684 \u985e\u3002 \u7684\u8907\u96dc\u5ea6\u8a2d\u5b9a\u70ba 0.5\u3002\u9019\u4e9b\u62bd\u53d6\u51fa\u4f86\u7684\u591a\u5b57\u8a5e\u6216\u8a5e\u7d44\u52a0\u5165\u8a5e\u5178\u5f8c\uff0c\u5c0d\u8ad6\u6587\u8cc7\u6599\u9032\u884c \u63a5\u8457\u5c07\u53d6\u51fa\u4f86\u7684\u8a5e\u8a9e\u9032\u884c\u7814\u7a76\u4e3b\u984c\u53e2\u805a\u3002\u9032\u884c\u8a5e\u8a9e\u7684 cliques \u53e2\u96c6\u6642\uff0c\u6211\u5011\u5206 \u7279\u5b9a\u7684\u7814\u7a76\u4e3b\u984c\u3002\u9664\u4e86\u53e2\u96c6 3\u3001\u53e2\u96c6 11 \u8207\u53e2\u96c6 29 \u7531\u610f\u7fa9\u8f03\u5ee3\u6cdb\u7684\u8a5e\u8a9e\u5f62\u6210\u4e4b\u5916\uff0c \u8868\u7684\u5e74\u4efd\u3002\u8868\u4e09\u53ef\u4ee5\u9a57\u8b49\u65e9\u671f\u7684\u8a08\u7b97\u8a9e\u8a00\u5b78\u591a\u4ee5\u898f\u5247\u5f0f\u7684\u8a9e\u6cd5\u6a21\u5f0f\u8207\u5256\u6790\u70ba\u4e3b\uff0c \u7d50\u679c\u521d\u6b65\u9a57\u8b49\u4e86\u672c\u8ad6\u6587\u6240\u63d0\u51fa\u65b9\u6cd5\u7684\u53ef\u884c\u6027\u3002\u5f9e\u7814\u7a76\u7d50\u679c\u4e2d\uff0c\u6211\u5011\u4e5f\u767c\u73fe\u8a08\u7b97\u8a9e</td></tr><tr><td>\u5206\u8a5e\uff0c\u4f9d\u64da\u7b2c\u4e09\u7bc0\u7684\u65b9\u6cd5\u5c0d\u6240\u6709\u8a5e\u8a9e\u9032\u884c\u7d71\u8a08\uff0c\u904e\u6ffe\u53bb\u4e0d\u91cd\u8981\u7684\u8a5e\u8a9e\u3002\u6700\u5f8c\u7684\u7d50 \u5225\u4ee5\u7b2c\u56db\u7bc0\u4e2d\u539f\u5148\u7684\u8a5e\u8a9e\u7279\u5fb5\u5411\u91cf\u8207\u7d93\u904e SVD \u8655\u7406\u7684\u7279\u5fb5\u5411\u91cf\uff0ck \u503c\u70ba 30\u300160 \u53ca \u5176\u9918\u53e2\u96c6\u7684\u8a5e\u8a9e\u9593\u90fd\u5177\u6709\u76f8\u95dc\u6027\uff0c\u800c\u4e14\u53ef\u4ee5\u7528\u4f86\u4ee3\u8868\u8a08\u7b97\u8a9e\u8a00\u5b78\u9818\u57df\u4e2d\u7684\u7279\u5b9a\u7814 \u8fd1\u4f86\u5247\u8f03\u591a\u767c\u5c55\u7d71\u8a08\u5f0f\u8a9e\u8a00\u6a21\u578b\uff0c\u800c\u65b7\u8a5e\u5247\u662f\u4e00\u76f4\u4ee5\u4f86\u570b\u5167\u8a08\u7b97\u8a9e\u8a00\u5b78\u9818\u57df\u76f8\u7576 \u8a00\u5b78\u7814\u7a76\u8207\u5be6\u52d9\u61c9\u7528\u6709\u5bc6\u5207\u7684\u95dc\u4fc2\uff0c\u62bd\u53d6\u51fa\u4f86\u7684\u8a5e\u8a9e\u53e2\u96c6\u4e2d\u6709\u8a31\u591a\u8207\u6a5f\u5668\u7ffb\u8b6f\u3001</td></tr><tr><td>1 e \u679c\u5171\u5f97\u5230 343 \u500b\u95dc\u9375\u8a5e\u8a9e\u3002\u7531\u65bc\u7bc7\u5e45\u6240\u9650\uff0c\u7121\u6cd5\u5c07\u6240\u6709\u7684\u8a5e\u8a9e\u4e00\u4e00\u5217\u51fa\uff0c\u6211\u5011\u5c07 t e e K = \u03c7 \uff0c\u6bcf\u4e00\u500b\u5143\u7d20\u4ee3\u8868\u4e00\u500b\u7279\u5b9a\u8a5e\u8a9e\u662f 120\uff0c\u9032\u884c\u76f8\u95dc\u7a0b\u5ea6\u4f30\u7b97\u3002\u5c07\u76f8\u95dc\u7a0b\u5ea6\u7684\u95be\u503c\u8a2d\u70ba 0.4\uff0c\u7d93\u904e cliques \u53e2\u96c6\u8207\u53e2\u96c6\u5408 \u7a76\u4e3b\u984c\u3002\u6bd4\u65b9\u8aaa\uff0c\u53e2\u96c6 7 \u70ba\u8a9e\u97f3\u8fa8\u8a8d\u7684\u76f8\u95dc\u8a5e\u8a9e\u3001\u53e2\u96c6 9 \u5247\u70ba\u6587\u4ef6\u5206\u985e\u7684\u76f8\u95dc\u8a5e \u91cd\u8996\u7684\u7368\u7279\u554f\u984c\u3002 \u8a9e\u97f3\u8655\u7406\u548c\u8cc7\u8a0a\u6aa2\u7d22\u76f8\u95dc\uff0c\u5728\u8a9e\u8a00\u7684\u8a08\u7b97\u6a21\u5f0f\u4e0a\uff0c\u8a9e\u6cd5\u6a21\u5f0f\u8207\u5256\u6790\u3001\u65b7\u8a5e\u548c\u8a9e\u8a00</td></tr></table>", |
| "text": "\u500b\u8a5e\u8a9e\u5305\u542b\u65bc\u9019\u500b\u53e2\u96c6\u4e2d\uff0c\u5247 e i \u7684\u503c\u70ba 1\uff1b \u5426\u5247\u82e5\u662f\u9019\u500b\u53e2\u96c6\u4e0d\u5305\u542b\u9019\u500b\u8a5e\u8a9e\uff0ce i \u7684\u503c\u70ba 0\u3002\u5f0f(6)\u6240\u5f97\u5230\u7684\u7d50\u679c R \u03a7 \u4e5f\u662f\u4e00\u500b A. P. Bishop, \"Document Structure and Digital Libraries: How Researchers Mobilize Information in Journal Articles\", Information Processing and Management, 35, p255-279. aspect, logic, temporal, tense 9 document, hierarchical, text categorization, \u5206\u985e, \u6587\u4ef6, \u6587\u4ef6\u5206\u985e, \u7279\u5fb5 \u8868\u9054, \u5256\u6790, \u683c\u4f4d, \u8a0a\u606f, \u52d5\u8a5e, \u7d50\u69cb, \u8a5e\u985e, \u6f22\u8a9e, \u8a9e\u6cd5, \u8a9e\u6cd5\u6a21\u5f0f, \u8a9e\u610f, \u6a21\u5f0f, \u95dc\u4fc2 24 adaptive, compression, scheme, \u82f1\u6587, \u8cc7\u6599, \u8abf\u6574, \u58d3\u7e2e 25 csmart, databases, document, indexing, information retrieval, retrieval, text retrieval, \u6aa2\u7d22 26 grammars, parser, parsing, sentence 27 continuous, large vocabulary, mandarin, speaker, speech, speech recognition, telephone", |
| "num": null |
| } |
| } |
| } |
| } |
