| { |
| "paper_id": "O15-1009", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T08:10:19.010328Z" |
| }, |
| "title": "\u57fa\u65bc\u8c9d\u6c0f\u5b9a\u7406\u81ea\u52d5\u5206\u6790\u8a9e\u6599\u5eab\u8207\u6a19\u5b9a\u6587\u6b65 * A Bayesian approach to determine move tags in corpus", |
| "authors": [ |
| { |
| "first": "Chiung-Wen", |
| "middle": [], |
| "last": "\u5f35\u74ca\u6587", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Chang", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Jia-Lien", |
| "middle": [], |
| "last": "Hsu", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Jason", |
| "middle": [ |
| "S" |
| ], |
| "last": "\u5f35\u4fca\u76db", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| } |
| ], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "\u6458\u8981 \u5229\u7528\u79d1\u6280\u5e6b\u52a9\u8a9e\u8a00\u5b78\u7fd2\uff0c\u662f\u4e00\u500b\u91cd\u8981\u7684\u7814\u7a76\u8b70\u984c\uff0c\u82f1\u6587\u662f\u73fe\u4eca\u4eba\u5011\u4e3b\u8981\u7684\u6e9d\u901a\u8a9e\u8a00\uff0c\u5c0d\u65bc \u975e\u82f1\u8a9e\u9ad4\u7cfb\u7684\u570b\u5bb6\uff0c\u5b78\u7fd2\u82f1\u8a9e (\u5f9e\u807d\u529b\u3001\u95b1\u8b80\u5230\u5beb\u4f5c) \u662f\u4e00\u4ef6\u56f0\u96e3\u7684\u4e8b\u60c5\u3002\u5c24\u5176\u5728\u5beb\u4f5c\u65b9\u9762\uff0c\u7531 \u65bc\u82f1\u6587\u6587\u6cd5\u8ddf\u4e2d\u6587\u6587\u6cd5\u4e0a\u7684\u5dee\u7570\uff0c\u5c0e\u81f4\u5728\u5b78\u7fd2\u82f1\u6587\u5beb\u4f5c\u6642\uff0c\u5e38\u5e38\u6703\u5c07\u7d44\u6210\u53e5\u5b50\u7684\u67b6\u69cb\u641e\u6df7\uff0c\u4f7f \u5f97\u5728\u5b78\u7fd2\u5beb\u4f5c\u6709\u8f03\u5927\u7684\u56f0\u96e3\u3002 \u82f1\u6587\u5b78\u8853\u8ad6\u6587\u5beb\u4f5c\uff0c\u4e0d\u540c\u65bc\u4e00\u822c\u6587\u7ae0\u5beb\u4f5c\uff0c\u901a\u5e38\u6709\u660e\u78ba\u7684\u67b6\u69cb\u8207\u6bb5\u843d\uff0c\u5982\u300c\u7c21\u4ecb\u300d \u3001 \u300c\u76f8\u95dc \u6587\u737b\u300d \u3001 \u300c\u65b9\u6cd5\u300d \u3001 \u300c\u7d50\u679c\u300d\u7b49\uff0c\u6b64\u7d50\u69cb\u7a31\u70ba\u300c\u6587\u6b65\u300d \u3002\u6b64\u5916\uff0c\u5b78\u8853\u8ad6\u6587\u5beb\u4f5c\u8207\u4e00\u822c\u5beb\u4f5c\u6709\u4e9b\u8a31\u7684\u4e0d \u540c\uff0c\u5728\u5beb\u4f5c\u7684\u7528\u8a5e\u4e0a\u5c31\u6709\u4e9b\u5dee\u7570\uff0c\u56e0\u6b64\uff0c\u70ba\u4e86\u5e6b\u52a9\u9700\u8981\u5beb\u5b78\u8853\u8ad6\u6587\u7684\u540c\u5b78\u5011\uff0c\u6211\u5011\u53c3\u8003\u5b78\u8853\u8ad6 \u6587\u7684\u6587\u6b65\u67b6\u69cb\uff0c\u8a2d\u8a08\u6587\u6b65\u5206\u985e\u5668\u8a13\u7df4\u8a9e\u8a00\u6a21\u7d44\uff0c\u64f7\u53d6\u5728\u7279\u5b9a\u6587\u6b65\u4f7f\u7528\u7684\u5b57\u8a5e\u3002 \u5728\u8a9e\u8a00\u8655\u7406\u65b9\u9762\uff0c\u5b78\u8005\u5011\u4f9d\u7167\u6587\u6b65\u67b6\u69cb\uff0c\u63d0\u51fa\u81ea\u52d5\u5316\u5206\u6790\uff0c\u4f46\u662f\u5728\u8a13\u7df4\u8a9e\u8a00\u6a21\u7d44\u4e2d\u901a\u5e38\u9700 \u8981\u5927\u91cf\u4eba\u5de5\u6a19\u8a3b\u8cc7\u6599\uff0c\u70ba\u4e86\u964d\u4f4e\u4eba\u5de5\u6a19\u8a3b\u7684\u90e8\u5206\uff0c\u6211\u5011\u5c07\u5c08\u5bb6\u6574\u7406\u6b78\u7d0d\u7684\u8a5e\u5f59\uff0c\u900f\u904e\u6a5f\u5668\u5b78\u7fd2 \u8207\u8fed\u4ee3 (bootstraping) \u7684\u65b9\u6cd5\u9054\u5230\u5b78\u7fd2\u6548\u679c\uff0c\u518d\u5229\u7528\u8a13\u7df4\u904e\u7684\u8a9e\u8a00\u6a21\u578b\uff0c\u9810\u6e2c\u6587\u7ae0\u53e5\u5b50\u7576\u4e2d\u7684 \u6587\u6b65\u3002 \u5728\u672c\u7814\u7a76\u4e2d\uff0c\u6211\u5011\u63d0\u51fa\u4e00\u5957\u7cfb\u7d71\uff0c\u4ee5\u8c9d\u6c0f\u65b9\u6cd5 (Bayesian approach) \u505a\u8a9e\u8a00\u6587\u6b65\u5206\u6790\uff0c\u6b64 \u7cfb\u7d71\u5206\u70ba\u5169\u90e8\u5206\uff0c\u4e00\u70ba\u8a13\u7df4\u968e\u6bb5 (Training phase)\uff0c\u53e6\u70ba\u6e2c\u8a66\u968e\u6bb5 (Testing phase)\u3002\u5728\u8a13\u7df4\u968e \u6bb5\u4e2d\uff0c\u900f\u904e\u5927\u91cf\u7684\u6587\u672c (Corpus) \u5efa\u7acb\u5b78\u7fd2\u6a21\u578b\uff0c\u63a1\u7528\u5c08\u9580\u8490\u96c6\u5b78\u8853\u8ad6\u6587\u7c21\u4ecb\u7684\u8a9e\u6599\u96c6 (Cite-SeerX) \u8207\u521d\u59cb\u898f\u5247 (Initial pattern) \u505a\u70ba\u5206\u6790\u7684\u4f9d\u64da\uff0c\u5229\u7528\u8c9d\u6c0f\u65b9\u6cd5\u5224\u65b7\u8a9e\u6599\u5eab\u4e2d\u6bcf\u7bc7\u7c21\u4ecb\u88e1 \u7684\u53e5\u5b50\u6240\u5c6c\u7684\u6587\u6b65 (move)\uff0c\u7576\u53e5\u5b50\u88ab\u6a19\u5b9a\u5b8c\u6587\u6b65\u4e4b\u5f8c\uff0c\u5229\u7528\u8fed\u4ee3\u7684\u65b9\u6cd5\u66f4\u65b0\u8c9d\u6c0f\u6a21\u578b\uff0c\u9054\u5230\u5b78 \u7fd2\u6548\u679c\u3002\u800c\u5728\u6e2c\u8a66\u6a21\u578b\u4e2d\uff0c\u5c07\u8a13\u7df4\u968e\u6bb5\u5f97\u5230\u7684\u7d50\u679c\uff0c\u7d66\u4e88\u4e00\u7bc7\u65b0\u7684\u7c21\u4ecb\uff0c\u4e00\u6a23\u900f\u904e\u8c9d\u6c0f\u65b9\u6cd5\uff0c \u9810\u6e2c\u6587\u6b65\uff0c\u7d93\u904e\u6e2c\u8a66\u968e\u6bb5\uff0c\u6211\u5011\u5f97\u5230\u6587\u6b65\u9810\u6e2c\u7cbe\u78ba\u7387\u70ba 56%\u3002 \u95dc\u9375\u8a5e\uff1a\u5b78\u8853\u82f1\u6587\u5beb\u4f5c\u3001\u8f14\u52a9\u5beb\u4f5c\u3001\u6587\u6b65\u5206\u6790", |
| "pdf_parse": { |
| "paper_id": "O15-1009", |
| "_pdf_hash": "", |
| "abstract": [ |
| { |
| "text": "\u6458\u8981 \u5229\u7528\u79d1\u6280\u5e6b\u52a9\u8a9e\u8a00\u5b78\u7fd2\uff0c\u662f\u4e00\u500b\u91cd\u8981\u7684\u7814\u7a76\u8b70\u984c\uff0c\u82f1\u6587\u662f\u73fe\u4eca\u4eba\u5011\u4e3b\u8981\u7684\u6e9d\u901a\u8a9e\u8a00\uff0c\u5c0d\u65bc \u975e\u82f1\u8a9e\u9ad4\u7cfb\u7684\u570b\u5bb6\uff0c\u5b78\u7fd2\u82f1\u8a9e (\u5f9e\u807d\u529b\u3001\u95b1\u8b80\u5230\u5beb\u4f5c) \u662f\u4e00\u4ef6\u56f0\u96e3\u7684\u4e8b\u60c5\u3002\u5c24\u5176\u5728\u5beb\u4f5c\u65b9\u9762\uff0c\u7531 \u65bc\u82f1\u6587\u6587\u6cd5\u8ddf\u4e2d\u6587\u6587\u6cd5\u4e0a\u7684\u5dee\u7570\uff0c\u5c0e\u81f4\u5728\u5b78\u7fd2\u82f1\u6587\u5beb\u4f5c\u6642\uff0c\u5e38\u5e38\u6703\u5c07\u7d44\u6210\u53e5\u5b50\u7684\u67b6\u69cb\u641e\u6df7\uff0c\u4f7f \u5f97\u5728\u5b78\u7fd2\u5beb\u4f5c\u6709\u8f03\u5927\u7684\u56f0\u96e3\u3002 \u82f1\u6587\u5b78\u8853\u8ad6\u6587\u5beb\u4f5c\uff0c\u4e0d\u540c\u65bc\u4e00\u822c\u6587\u7ae0\u5beb\u4f5c\uff0c\u901a\u5e38\u6709\u660e\u78ba\u7684\u67b6\u69cb\u8207\u6bb5\u843d\uff0c\u5982\u300c\u7c21\u4ecb\u300d \u3001 \u300c\u76f8\u95dc \u6587\u737b\u300d \u3001 \u300c\u65b9\u6cd5\u300d \u3001 \u300c\u7d50\u679c\u300d\u7b49\uff0c\u6b64\u7d50\u69cb\u7a31\u70ba\u300c\u6587\u6b65\u300d \u3002\u6b64\u5916\uff0c\u5b78\u8853\u8ad6\u6587\u5beb\u4f5c\u8207\u4e00\u822c\u5beb\u4f5c\u6709\u4e9b\u8a31\u7684\u4e0d \u540c\uff0c\u5728\u5beb\u4f5c\u7684\u7528\u8a5e\u4e0a\u5c31\u6709\u4e9b\u5dee\u7570\uff0c\u56e0\u6b64\uff0c\u70ba\u4e86\u5e6b\u52a9\u9700\u8981\u5beb\u5b78\u8853\u8ad6\u6587\u7684\u540c\u5b78\u5011\uff0c\u6211\u5011\u53c3\u8003\u5b78\u8853\u8ad6 \u6587\u7684\u6587\u6b65\u67b6\u69cb\uff0c\u8a2d\u8a08\u6587\u6b65\u5206\u985e\u5668\u8a13\u7df4\u8a9e\u8a00\u6a21\u7d44\uff0c\u64f7\u53d6\u5728\u7279\u5b9a\u6587\u6b65\u4f7f\u7528\u7684\u5b57\u8a5e\u3002 \u5728\u8a9e\u8a00\u8655\u7406\u65b9\u9762\uff0c\u5b78\u8005\u5011\u4f9d\u7167\u6587\u6b65\u67b6\u69cb\uff0c\u63d0\u51fa\u81ea\u52d5\u5316\u5206\u6790\uff0c\u4f46\u662f\u5728\u8a13\u7df4\u8a9e\u8a00\u6a21\u7d44\u4e2d\u901a\u5e38\u9700 \u8981\u5927\u91cf\u4eba\u5de5\u6a19\u8a3b\u8cc7\u6599\uff0c\u70ba\u4e86\u964d\u4f4e\u4eba\u5de5\u6a19\u8a3b\u7684\u90e8\u5206\uff0c\u6211\u5011\u5c07\u5c08\u5bb6\u6574\u7406\u6b78\u7d0d\u7684\u8a5e\u5f59\uff0c\u900f\u904e\u6a5f\u5668\u5b78\u7fd2 \u8207\u8fed\u4ee3 (bootstraping) \u7684\u65b9\u6cd5\u9054\u5230\u5b78\u7fd2\u6548\u679c\uff0c\u518d\u5229\u7528\u8a13\u7df4\u904e\u7684\u8a9e\u8a00\u6a21\u578b\uff0c\u9810\u6e2c\u6587\u7ae0\u53e5\u5b50\u7576\u4e2d\u7684 \u6587\u6b65\u3002 \u5728\u672c\u7814\u7a76\u4e2d\uff0c\u6211\u5011\u63d0\u51fa\u4e00\u5957\u7cfb\u7d71\uff0c\u4ee5\u8c9d\u6c0f\u65b9\u6cd5 (Bayesian approach) \u505a\u8a9e\u8a00\u6587\u6b65\u5206\u6790\uff0c\u6b64 \u7cfb\u7d71\u5206\u70ba\u5169\u90e8\u5206\uff0c\u4e00\u70ba\u8a13\u7df4\u968e\u6bb5 (Training phase)\uff0c\u53e6\u70ba\u6e2c\u8a66\u968e\u6bb5 (Testing phase)\u3002\u5728\u8a13\u7df4\u968e \u6bb5\u4e2d\uff0c\u900f\u904e\u5927\u91cf\u7684\u6587\u672c (Corpus) \u5efa\u7acb\u5b78\u7fd2\u6a21\u578b\uff0c\u63a1\u7528\u5c08\u9580\u8490\u96c6\u5b78\u8853\u8ad6\u6587\u7c21\u4ecb\u7684\u8a9e\u6599\u96c6 (Cite-SeerX) \u8207\u521d\u59cb\u898f\u5247 (Initial pattern) \u505a\u70ba\u5206\u6790\u7684\u4f9d\u64da\uff0c\u5229\u7528\u8c9d\u6c0f\u65b9\u6cd5\u5224\u65b7\u8a9e\u6599\u5eab\u4e2d\u6bcf\u7bc7\u7c21\u4ecb\u88e1 \u7684\u53e5\u5b50\u6240\u5c6c\u7684\u6587\u6b65 (move)\uff0c\u7576\u53e5\u5b50\u88ab\u6a19\u5b9a\u5b8c\u6587\u6b65\u4e4b\u5f8c\uff0c\u5229\u7528\u8fed\u4ee3\u7684\u65b9\u6cd5\u66f4\u65b0\u8c9d\u6c0f\u6a21\u578b\uff0c\u9054\u5230\u5b78 \u7fd2\u6548\u679c\u3002\u800c\u5728\u6e2c\u8a66\u6a21\u578b\u4e2d\uff0c\u5c07\u8a13\u7df4\u968e\u6bb5\u5f97\u5230\u7684\u7d50\u679c\uff0c\u7d66\u4e88\u4e00\u7bc7\u65b0\u7684\u7c21\u4ecb\uff0c\u4e00\u6a23\u900f\u904e\u8c9d\u6c0f\u65b9\u6cd5\uff0c \u9810\u6e2c\u6587\u6b65\uff0c\u7d93\u904e\u6e2c\u8a66\u968e\u6bb5\uff0c\u6211\u5011\u5f97\u5230\u6587\u6b65\u9810\u6e2c\u7cbe\u78ba\u7387\u70ba 56%\u3002 \u95dc\u9375\u8a5e\uff1a\u5b78\u8853\u82f1\u6587\u5beb\u4f5c\u3001\u8f14\u52a9\u5beb\u4f5c\u3001\u6587\u6b65\u5206\u6790", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Abstract", |
| "sec_num": null |
| } |
| ], |
| "body_text": [ |
| { |
| "text": "follow a simple and succinct picture of the organizational patterns, called move. This paper introduces a method for computational analysis of move structures, the Background-Purpose-Method-Result-Conclusion in this paper, in abstracts and introductions of research documents, instead of manually time-consuming and labor-intensive analysis process. In our approach, sentences in a given abstract and introduction are automatically analyzed and labeled with a specific move (i.e., B-P-M-R-C in this paper) to reveal various rhetorical functions. As a result, it is expected that the automatic analytical tool for move structures will facilitate non-native speakers or novice writers to be aware of appropriate move structures and internalize relevant knowledge to improve their writing.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "In this paper, we propose a Bayesian approach to determine move tags for research articles. The approach consists of two phases, training phase and testing phase. In the training phase, we build a Bayesian model based on a couples of given initial patterns and the corpus, a subset of CiteSeerX. In the beginning, the priori probability of Bayesian model solely relies on initial patterns. Subsequently, with respect to the corpus, we process each document one by one: extract features, determine tags, and update the Bayesian model iteratively. In the testing phase, we compare our results with tags which are manually assigned by the experts. In our experiments, the promising accuracy of the proposed approach reaches 56%. ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "EQUATION", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [ |
| { |
| "start": 0, |
| "end": 8, |
| "text": "EQUATION", |
| "ref_id": "EQREF", |
| "raw_str": "B 1 P 1 M 1 R 1 C 1 ng 2 a basic issue for B 2 P 2 M 2 R 2 C 2 \u2026 \u2026 \u2026 \u2026 \u2026 \u2026 \u2026 ng m role of the presequence B m P m M m R m C m P (ng 1 ) = B 1 + P 1 + M 1 + R 1 + C 1 \u2211 m i=1 (B i + P i + M i + R i + C i )", |
| "eq_num": "(6)" |
| } |
| ], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "\u5247\u6703\u5224\u65b7\u6b64 ng 1 \u662f\u5426\u5b58\u5728\u65bc\u521d\u59cb\u8868 (\u8868 3)\uff0c\u82e5\u5b58\u5728\u65bc\u521d\u59cb\u8868 (CT )\uff0c\u5247\u6703\u8a08\u7b97 N-\u9023\u8a5e\u5728\u8a72 \u6587\u6b65 (B) \u6b21\u6578\u51fa\u73fe\u7684\u6a5f\u7387\u3002", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "P (ng 1 |B) = B 1 \u2211 m i=1 B i (7)", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "\u82e5\u8a72 N-\u9023\u8a5e\u4e0d\u5b58\u5728\u65bc\u521d\u59cb\u8868\u683c\u6216\u662f\u5728\u672a\u66fe\u51fa\u73fe\u65bc\u67d0\u6587\u6b65\uff0c\u6bd4\u5982\"a basic issue for\" \u6b64 N-\u9023 \u8a5e\u4e0d\u66fe\u51fa\u73fe\u65bc\u7d50\u8ad6 (C)\uff0c\u5247\u6703\u7d66\u4e88\u6975\u5c0f\u503c (\u03b4 = 10 \u22128 ) \u7576\u4f5c\u6a5f\u7387\u3002\u5c07\u6bcf\u500b N-\u9023\u8a5e\uff0c\u5c0d\u7167\u521d\u59cb\u898f \u5247\u8868 (CT )\uff0c\u56e0\u6b64 S 1 \u900f\u904e\u904b\u7b97\u5247\u6703\u5f97\u8fd1\u4f3c\u6240\u5c6c\u7684\u6587\u6b65\u6a5f\u7387\u503c\u3002\u800c\u5404\u6587\u6b65\u7684\u6a5f\u7387\uff0c\u5247\u662f\u4f9d\u7167\u6587\u6b65 \u6b21\u6578\u505a\u70ba\u4f9d\u64da\u3002 \u7531\u65bc\u53e5\u5b50\u7d44\u6210\u55ae\u5b57\u7684\u591a\u5be1\uff0c\u6703\u5f71\u97ff\u8a08\u7b97\u4e0a\u7684\u516c\u5e73\u6027\uff0c\u6240\u4ee5\u6211\u5011\u5c07\u7d50\u679c\u6b63\u898f\u5316\u3002", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "normalized (P (B|S 1 )) = P (B|S 1 ) # of n-gram in S 1 (8) 3.4.3 \u6587\u6b65\u6a19\u5b9a \u5728\u6587\u6b65\u6a19\u5b9a\u4e0a\uff0c\u5728\u672c\u6587\u4e2d\uff0c\u5148\u5c07\u4e00\u7bc7\u6587\u7ae0 (D 1 ) \u4e2d\u6240\u6709\u53e5\u5b50\u7684\u5404\u6587\u6b65\u6a5f\u7387\u8a08\u7b97\u5b8c\u7562\uff0c\u624d\u9010\u53e5\u6a19 \u5b9a\u6587\u6b65\u3002 \u800c\u6bcf\u500b\u6587\u6b65\u8981\u6a19\u5b9a\u5e7e\u500b\u53e5\u5b50\uff0c\u5247\u662f\u4f9d\u64da\u7d66\u5b9a\u7684\u6bd4\u4f8b\u53bb\u505a\u8a08\u7b97\uff0c\u7531\u65bc\u53e5\u6578\u4e0d\u80fd\u70ba\u5c0f\u6578\uff0c\u6240\u4ee5 \u53d6\u56db\u6368\u4e94\u5165\u7684\u65b9\u6cd5\u3002 \u8868 6: \u4e00\u7bc7\u6587\u7ae0\u4e2d (D 1 )\uff0c\u6587\u6b65\u53e5\u6578\u7b97\u6cd5\u3002 move-tag \u6587\u7ae0\u53e5\u6578\u6bd4\u4f8b \u53e5\u6578 B 0.15 \u00d7 6 = 0.9 1 P 0.20 \u00d7 6 = 1.2 1 M 0.30 \u00d7 6 = 1.8 2 C 0.15 \u00d7 6 = 0.9 1 R 6 \u2212 (1 + 1 + 2 + 1) 1 \u70ba\u4e86\u907f\u514d\u67d0\u4e00\u6587\u6b65\u9020\u6210\u591a\u6578\u5236 (Majority rule) \u7d50\u679c\uff0c\u6211\u5011\u6839\u64da\u6587\u6b65\u5728 Corpus \u5167\u6587\u5beb\u7684\u6bd4 \u4f8b\u591a\u5be1\uff0c\u4f9d\u5e8f\u6a19\u5b9a\u6587\u6b65 (\u4ee5\u8868 6\u70ba\u4f8b\uff0c\u5148\u6a19\u5b9a B \u5f80\u5f8c\u9806\u5e8f\u70ba C \u2192 P \u2192 R \u2192 M )\u3002\u7531\u65bc\u6211\u5011\u662f \u7531\u4e00\u7bc7\u6587\u7ae0\u5224\u5b9a\u53e5\u5b50\u6587\u6b65\uff0c\u4f9d\u7167\u6bd4\u4f8b\uff0c\u6211\u5011\u5148\u6a19\u5b9a\u70ba B \u7684\u53e5\u5b50\u3002 \u2235 B 2 = max{B 1 , B 2 , ..., B 6 } \u2234 S 2 \u2190 B (9) \u82e5\u8a72\u53e5 (S 2 ) \u5df2\u7d93\u88ab\u6a19\u4e0a\u6a19\u7c64 (B)\uff0c\u5247\u5c07\u53e5\u5b50\u79fb\u9664\u5e8f\u5217\u4e2d\uff0c\u7d93\u7531\u8868 6\u8a08\u7b97\uff0c\u6587\u7ae0\u7576\u4e2d\u70ba B \u7684 \u5167\u5bb9\u70ba\u4e00\u53e5\uff0c\u5247\u63db\u6a19\u5b9a\u4e0b\u4e00\u500b\u6587\u6b65 C\u3002 \u8868 7: \u7d93\u904e\u7b2c\u4e00\u6b21\u6587\u6b65\u6a19\u5b9a Sentence B P M R C S 1 B 1 P 1 M 1 R 1 C 1 S 2 B 1 P 1 M 1 R 1 C 1 S 3 B 3 P 3 M 3 R 3 C 3 S 4 B 4 P 4 M 4 R 4 C 4 S 5 B 5 P 5 M 5 R 5 C 5 S 6 B 6 P 6 M 6 R 6 C 6 \u53cd\u8986\u6a19\u5b9a\u904e\u7a0b\uff0c\u5c07\u6587\u7ae0\u7576\u4e2d\u7684\u53e5\u5b50\u6a19\u4e0a\u6587\u6b65\u3002 \u2235 C 6 = max{C 1 , C 3 , ..., C 6 } \u2234 S 6 \u2190 C (10) \u8868 8: \u7d93\u904e\u7b2c\u4e8c\u6b21\u6a19\u5b9a Sentence B P M R C S 1 B 1 P 1 M 1 R 1 C 1 S 2 B 2 P 2 M 2 R 2 C 2 S 3 B 3 P 3 M 3 R 3 C 3 S 4 B 4 P 4 M 4 R 4 C 4 S 5 B 5 P 5 M 5 R 5 C 5 S 6 B 6 P 6 M 6 R 6 C 6", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "\u7576\u4e00\u7bc7\u6587\u7ae0\u7576\u4e2d\u6240\u5305\u542b\u7684\u53e5\u5b50\u90fd\u5df2\u7d93\u6a19\u4e0a\u6587\u6b65\uff0c\u800c\u6211\u5011\u4e5f\u6703\u6839\u64da\u7d50\u679c\uff0c\u66f4\u65b0 CT \u76f8\u5c0d\u61c9 \u7684\u898f\u5247 (\u8868 3)\u3002\u5047\u8a2d S 1 \u88ab\u6a19\u5b9a\u70ba B\uff0c\u800c\u53e5\u5b50\u7576\u4e2d\u5305\u542b\u4e00\u500b N-\u9023\u8a5e\"ng: glyoxysomal citrate synthase in\"\uff0c\u4e0d\u5b58\u5728 CT \uff0c\u4f9d\u7167\u53e5\u5b50\u88ab\u6a19\u5b9a\u7684\u6587\u6b65\uff0c\u65b0\u589e ng \u81f3 CT \u4e2d\u4e26\u5728 B \u7d66\u4e88\u521d\u59cb\u6b21\u6578 (\u8868 9)\u3002 ", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "CiteSeerx: http://citeseerx.ist.psu.edu/about/site", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| } |
| ], |
| "back_matter": [], |
| "bib_entries": { |
| "BIBREF0": { |
| "ref_id": "b0", |
| "title": "Teaching EFL students to extract structural information from abstracts", |
| "authors": [ |
| { |
| "first": "N", |
| "middle": [], |
| "last": "Graetz", |
| "suffix": "" |
| } |
| ], |
| "year": 1985, |
| "venue": "Readings for Professional Purposes: Methods and Materials in Teaching Languages", |
| "volume": "", |
| "issue": "", |
| "pages": "225--335", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "N. Graetz, \"Teaching EFL students to extract structural information from abstracts,\" in Readings for Professional Purposes: Methods and Materials in Teaching Languages, J. M. Kline and A. K. Pugh, Eds., 1985, pp. 225 -335.", |
| "links": null |
| }, |
| "BIBREF1": { |
| "ref_id": "b1", |
| "title": "Discoursal flaws in medical english abstracts: A genre analysis per research-and text-type", |
| "authors": [ |
| { |
| "first": "F", |
| "middle": [], |
| "last": "Salager-Meyer", |
| "suffix": "" |
| } |
| ], |
| "year": 1990, |
| "venue": "Text -Interdisciplinary Journal for the Study of Discourse", |
| "volume": "10", |
| "issue": "4", |
| "pages": "365--384", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "F. Salager-Meyer, \"Discoursal flaws in medical english abstracts: A genre analysis per research-and text-type,\" Text -Interdisciplinary Journal for the Study of Discourse, vol. 10, no. 4, pp. 365-384, 1990.", |
| "links": null |
| }, |
| "BIBREF2": { |
| "ref_id": "b2", |
| "title": "Using sectioning information for text retrieval: a case study with the medline abstracts", |
| "authors": [ |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Shimbo", |
| "suffix": "" |
| }, |
| { |
| "first": "T", |
| "middle": [], |
| "last": "Yamasaki", |
| "suffix": "" |
| }, |
| { |
| "first": "Y", |
| "middle": [], |
| "last": "Matsumoto", |
| "suffix": "" |
| } |
| ], |
| "year": 2003, |
| "venue": "Proceedings of Second International Workshop on Active Mining (AM'03)", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "M. Shimbo, T. Yamasaki, and Y. Matsumoto, \"Using sectioning information for text re- trieval: a case study with the medline abstracts,\" in Proceedings of Second International Workshop on Active Mining (AM'03), 2003.", |
| "links": null |
| }, |
| "BIBREF3": { |
| "ref_id": "b3", |
| "title": "American national standard for writing abstracts, ser", |
| "authors": [], |
| "year": 1997, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "39--53", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "American National Standards Institute, American national standard for writing abstracts, ser. Z39-14. Bethesda, Maryland, USA: NISO Press, 1997.", |
| "links": null |
| }, |
| "BIBREF4": { |
| "ref_id": "b4", |
| "title": "Genre analysis: English in academic and research settings", |
| "authors": [ |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Swales", |
| "suffix": "" |
| } |
| ], |
| "year": 1990, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "J. Swales, Genre analysis: English in academic and research settings. Cambridge Univer- sity Press, 1990.", |
| "links": null |
| }, |
| "BIBREF5": { |
| "ref_id": "b5", |
| "title": "Science research writing: For non-native speakers of English", |
| "authors": [ |
| { |
| "first": "H", |
| "middle": [], |
| "last": "Glasman-Deal", |
| "suffix": "" |
| } |
| ], |
| "year": 2009, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "H. Glasman-Deal, Science research writing: For non-native speakers of English. Imperial College Press, 2009.", |
| "links": null |
| }, |
| "BIBREF6": { |
| "ref_id": "b6", |
| "title": "Writing up research", |
| "authors": [ |
| { |
| "first": "R", |
| "middle": [], |
| "last": "Weissberg", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [], |
| "last": "Buker", |
| "suffix": "" |
| } |
| ], |
| "year": 1990, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "R. Weissberg and S. Buker, Writing up research. Englewood Cliffs, NJ, USA: Prentice Hall, 1990.", |
| "links": null |
| }, |
| "BIBREF7": { |
| "ref_id": "b7", |
| "title": "A genre analysis of english and spanish research paper abstracts in experimental social sciences", |
| "authors": [ |
| { |
| "first": "P", |
| "middle": [ |
| "M" |
| ], |
| "last": "Mart\u0131\u0144", |
| "suffix": "" |
| } |
| ], |
| "year": 2003, |
| "venue": "English for Specific Purposes", |
| "volume": "22", |
| "issue": "1", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "P. M. Mart\u0131\u0144, \"A genre analysis of english and spanish research paper abstracts in experi- mental social sciences,\" English for Specific Purposes, vol. 22, no. 1, pp. 25 -43, 2003. \u8868 15: \u7d93\u8a13\u7df4\u589e\u52a0\u898f\u5247\u7684 CT \u8868 \u7bc7\u6578 1,000 2,000 3,000 4,000 5,000", |
| "links": null |
| }, |
| "BIBREF10": { |
| "ref_id": "b10", |
| "title": "Summarizing scientific articles: Experiments with relevance and rhetorical status", |
| "authors": [ |
| { |
| "first": "S", |
| "middle": [], |
| "last": "Teufel", |
| "suffix": "" |
| }, |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Moens", |
| "suffix": "" |
| } |
| ], |
| "year": 2002, |
| "venue": "Comput. Linguist", |
| "volume": "28", |
| "issue": "4", |
| "pages": "409--445", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "S. Teufel and M. Moens, \"Summarizing scientific articles: Experiments with relevance and rhetorical status,\" Comput. Linguist., vol. 28, no. 4, pp. 409-445, 2002. [Online].", |
| "links": null |
| }, |
| "BIBREF11": { |
| "ref_id": "b11", |
| "title": "Available", |
| "authors": [], |
| "year": null, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": { |
| "DOI": [ |
| "10.1162/089120102762671936" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Available: http://dx.doi.org/10.1162/089120102762671936", |
| "links": null |
| }, |
| "BIBREF12": { |
| "ref_id": "b12", |
| "title": "USTC system for Blizzard Challenge 2006: an improved hmm-based speech synthesis method", |
| "authors": [ |
| { |
| "first": "Z.-H", |
| "middle": [], |
| "last": "Ling", |
| "suffix": "" |
| }, |
| { |
| "first": "Y.-J", |
| "middle": [], |
| "last": "Wu", |
| "suffix": "" |
| }, |
| { |
| "first": "Y.-P", |
| "middle": [], |
| "last": "Wang", |
| "suffix": "" |
| }, |
| { |
| "first": "L", |
| "middle": [], |
| "last": "Qin", |
| "suffix": "" |
| }, |
| { |
| "first": "R.-H", |
| "middle": [], |
| "last": "Wang", |
| "suffix": "" |
| } |
| ], |
| "year": 2006, |
| "venue": "Proceedings of Blizzard Challenge Workshop", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Z.-H. Ling, Y.-J. Wu, Y.-P. Wang, L. Qin, and R.-H. Wang, \"USTC system for Blizzard Challenge 2006: an improved hmm-based speech synthesis method,\" in Proceedings of Blizzard Challenge Workshop, 2006.", |
| "links": null |
| }, |
| "BIBREF13": { |
| "ref_id": "b13", |
| "title": "Computational analysis of move structures in academic abstracts", |
| "authors": [ |
| { |
| "first": "J.-C", |
| "middle": [], |
| "last": "Wu", |
| "suffix": "" |
| }, |
| { |
| "first": "Y.-C", |
| "middle": [], |
| "last": "Chang", |
| "suffix": "" |
| }, |
| { |
| "first": "H.-C", |
| "middle": [], |
| "last": "Liou", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [ |
| "S" |
| ], |
| "last": "Chang", |
| "suffix": "" |
| } |
| ], |
| "year": 2006, |
| "venue": "Proceedings of the COLING/ACL on Interactive Presentation Sessions, ser. COLING-ACL '06", |
| "volume": "", |
| "issue": "", |
| "pages": "41--44", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "J.-C. Wu, Y.-C. Chang, H.-C. Liou, and J. S. Chang, \"Computational analysis of move structures in academic abstracts,\" in Proceedings of the COLING/ACL on Interactive Presentation Sessions, ser. COLING-ACL '06. Stroudsburg, PA, USA: Association for Computational Linguistics, 2006, pp. 41-44.", |
| "links": null |
| }, |
| "BIBREF14": { |
| "ref_id": "b14", |
| "title": "A sentence classification system for multi biomedical literature summarization", |
| "authors": [ |
| { |
| "first": "Y", |
| "middle": [], |
| "last": "Yamamoto", |
| "suffix": "" |
| }, |
| { |
| "first": "T", |
| "middle": [], |
| "last": "Takagi", |
| "suffix": "" |
| } |
| ], |
| "year": 2005, |
| "venue": "Proceedings of 21st International Conference on Data Engineering Workshops", |
| "volume": "", |
| "issue": "", |
| "pages": "1163--1163", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Y. Yamamoto and T. Takagi, \"A sentence classification system for multi biomedical litera- ture summarization,\" in Proceedings of 21st International Conference on Data Engineering Workshops. IEEE, 2005, pp. 1163-1163.", |
| "links": null |
| }, |
| "BIBREF15": { |
| "ref_id": "b15", |
| "title": "Bayesian semi-supervised chinese word segmentation for statistical machine translation", |
| "authors": [ |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Xu", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Gao", |
| "suffix": "" |
| }, |
| { |
| "first": "K", |
| "middle": [], |
| "last": "Toutanova", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [], |
| "last": "Ney", |
| "suffix": "" |
| } |
| ], |
| "year": 2008, |
| "venue": "Proceedings of the 22nd International Conference on Computational Linguistics", |
| "volume": "1", |
| "issue": "", |
| "pages": "1017--1024", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "J. Xu, J. Gao, K. Toutanova, and H. Ney, \"Bayesian semi-supervised chinese word segmentation for statistical machine translation,\" in Proceedings of the 22nd International Conference on Computational Linguistics -Volume 1, ser. COLING '08. Stroudsburg, PA, USA: Association for Computational Linguistics, 2008, pp. 1017-1024. [Online].", |
| "links": null |
| }, |
| "BIBREF16": { |
| "ref_id": "b16", |
| "title": "Available", |
| "authors": [], |
| "year": null, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Available: http://dl.acm.org/citation.cfm?id=1599081.1599209", |
| "links": null |
| } |
| }, |
| "ref_entries": { |
| "TABREF0": { |
| "text": "\u9023\u8a5e (S = {ng 1 , ng 2 , ... }) \u56de\u994b\u5230\u521d\u59cb\u8868\uff0c\u85c9\u4ee5\u9054\u5230\u8a13 \u7df4\u7684\u6548\u679c\u3002 \u5728\u6e2c\u8a66\u968e\u6bb5\uff0c\u5247\u6703\u9078\u53d6\u4e00\u7bc7\u65b0\u7684\u6587\u7ae0\u7c21\u4ecb\uff0c\u900f\u904e\u8a13\u7df4\u5b8c\u7562\u7684\u6a21\u578b\u7d50\u679c\uff0c\u8a55\u4f30\u6587\u6b65\u6a19\u8a3b\u7684\u7cbe \u78ba\u7387\u3002 \u5728\u6b64\u7ae0\u7bc0\uff0c\u6558\u8ff0\u6211\u5011\u6240\u4f7f\u7528\u7684\u6f14\u7b97\u6cd5\uff0c\u5305\u542b\u8c9d\u6c0f\u5b9a\u7406\u6240\u9700\u8981\u7684\u5148\u9a57\u6a5f\u7387\u8207\u6587\u6b65\u7684\u6311\u9078\uff0c\u4e26 \u554f\u984c\u9673\u8ff0 \u7d66\u5b9a\uff1a\u4ee5\u5b78\u8853\u6587\u7ae0\u7d44\u6210\u7684\u8a9e\u6599\u96c6 (Corpus) \u8207\u521d\u59cb\u8a13\u7df4\u898f\u5247 (Initial pattern)\u6211\u5011\u5148\u8a08\u7b97\u4e00\u500b\u521d\u59cb\u6a21\u578b \u7d66\u5b9a\uff1a\u4e00\u7bc7\u5b78\u8853\u6587\u7ae0 D (D \u2208 Corpus) \u76ee\u6a19\uff1a\u70ba\u55ae\u4e00\u7bc7\u6587\u7ae0 (D = {S 1 , S 2 , ... }) \u4e2d\u6bcf\u4e00\u53e5\u5b50 S i \u5224\u5b9a\u53e5\u5b50\u6587\u6b65 \u6a19\u4e0a\u6587\u6b65\u6a19\u7c64 (move-tag = {B, P , M , R, C}) Speech, POS)\u3001\u610f\u5143\u96c6\u7d44 (Chunk)\u3002\u4f8b\u5982\u4e00\u7bc7\u6587\u7ae0\u4e2d\u5176\u4e2d\u4e00 \u53e5 (S 1 ) \u70ba\"Glyoxysomal citrate synthase in pumpkin is synthesized as a precursor that has one cleavable presequence at its N-terminal end.\" \u7d93\u904e enia tagger \u5206\u6790\u4e4b\u5f8c (\u8868 1)\uff0c\u6211\u5011\u4f9d\u7167\u7d50 \u679c\uff0c\u5c07\u53e5\u5b50\u6574\u7406\u6210\u4e09\u7a2e\u8868\u9054\u65b9\u5f0f\uff0c\u5206\u5225\u70ba1. \u539f\u59cb\u8cc7\u6599 (OW: Original word)\u5c07\u53e5\u5b50\u4fdd\u7559\u539f\u59cb\u8cc7\u6599\uff0c\u5305\u542b\u904e\u53bb\u5f0f\u3001\u8907\u6578\u7b49\u7b49\uff0c\u4f46\u662f\u6368\u53bb\u90e8\u5206\u7b26\u865f\uff0c\u4f7f\u5f97\u53e5\u5b50\u53ea\u7531\u55ae\u5b57 \u7d44\u6210\uff0c\u6240\u4ee5\u539f\u59cb\u53e5\u5b50 (S 1 )\uff0c\u5c07\u6703\u8f49\u6210\u5982\u4e0b\uff1a \"Glyoxysomal citrate synthase in pumpkin is synthesized as a precursor that has one cleavable presequence at its N-terminal end.\" \u5047\u8a2d\u70ba\u4e00\u7d44\u7368\u7acb N-\u9023\u8a5e\u6240\u7d44\u6210 (S 1 is approimated by set of n-grams as follows:{ng 1 , ng 2 , ...}) \u6240\u4ee5\u8981\u8a08\u7b97\u53e5\u5b50\u7684\u6240\u5c6c\u7684\u6587\u6b65\u6a5f \u7387\uff0c\u5728\u6b64\u5c07\u53e5\u5b50\u5283\u5206\u70ba N-\u9023\u8a5e\u4f86\u505a\u904b\u7b97\uff0c\u4f8b\u5982 S 1 \u8fd1\u4f3c\u70ba n \u500b N-\u9023\u8a5e\u6240\u7d44\u6210\u3002 S 1 \u2190 {ng 1 , ng 2 , ..., ng n } \u7bc4\u4f8b\u6587\u7ae0 D 1 = {S 1 , S 2 , ... , S 6 } Glyoxysomal citrate synthase in pumpkin is synthesized as one \u2026. S 2 To investigate the role of the presequence in the \u2026. S 3 Lmmunogold labeling and cell fractionation studies \u2026. S 4 The chimeric protein was transported to functionally \u2026. S 5 These observations indicated that the transport of \u2026. S 6 Site-directed mutagenesis of the conserved amino acids in \u2026.", |
| "html": null, |
| "content": "<table><tr><td>\u4e8c\u3001\u76f8\u95dc\u7814\u7a76 \u96a8\u8457\u8cc7\u8a0a\u767c\u5c55\uff0c\u70ba\u4e86\u8b93\u8cc7\u8a0a\u4ea4\u6d41\u5feb\u901f\uff0c\u95dc\u65bc\u81ea\u7136\u8a9e\u8a00\u8655\u7406\u70ba\u76f8\u7576\u91cd\u8981\u7684\u7814\u7a76\u9818\u57df\uff0c\u5728\u7d14\u6587 \u8868 2: \u5f9e Glasman-Deal \u64b0\u5beb\u7684\u66f8 [6] \u6240\u64f7\u53d6\u51fa\u90e8\u5206\u7684\u521d\u59cb\u898f\u5247 (Initial pattern) \u8868 4: \u521d\u59cb\u898f\u5247 (Initial pattern) \u4e2d\uff0c\u5404\u7a2e\u6587\u6b65\u7684\u6b21\u6578\u5206\u4f48</td></tr><tr><td>\u5b57\u7684\u61c9\u7528\u5305\u62ec\u6a5f\u5668\u7ffb\u8b6f\u3001\u62fc\u5b57\u6821\u6b63\u3001\u8cc7\u6599\u6aa2\u7d22\u7b49\u7b49\u3002\u8fd1\u5e74\u4f86\u5b78\u8005\u5c0d\u65bc\u5b78\u8853\u8ad6\u6587\u6216\u662f\u671f\u520a\uff0c\u6709\u9032 \u4e00\u6b65\u7684\u7814\u7a76 (Swales & Feak, 2004)\u3002\u4e3b\u8981\u91dd\u5c0d\u8ad6\u6587\u7684\u6bb5\u843d\u8207\u53e5\u5b50\u9032\u884c\u4eba\u70ba\u7684\u5206\u6790\u7814\u7a76\uff0c\u7d93\u904e\u6b78 Pattern move-tag B P M R C</td></tr><tr><td>\u7d0d\u4e4b\u5f8c\u63d0\u51fa\u95dc\u65bc\u8ad6\u6587\u4fee\u8fad\u7684\u67b6\u69cb\u898f\u5247-\u300c\u6587\u6b65\u300d \u3002\u5728\u672c\u7814\u7a76\u4e2d\uff0c\u5247\u662f\u91dd\u5c0d\u8ad6\u6587\u7684\u300c\u7c21\u4ecb\u300d\u9019\u4e00\u500b a basic issue for \u6b21\u6578 25 34 33 41 22</td></tr><tr><td>\u7ae0\u7bc0\u505a\u5206\u6790\uff0c\u63d0\u51fa\u81ea\u52d5\u5316\u5206\u6790\u8ad6\u6587\u6587\u6b65\u7d50\u69cb\u7684\u65b9\u6cd5\u3002 approach was developed by</td></tr><tr><td>\u5927\u90e8\u5206\u8ad6\u6587\u7c21\u4ecb\u6709\u8457\u7c21\u55ae\u6587\u6b65\u7d50\u69cb-IMRD [8]\uff0c\u5373\u70ba\u4ecb\u7d39 (Introduction)\u3001\u65b9\u6cd5 (Method)\u3001 majority of the tests</td></tr><tr><td>Keyword: Academic English Writing, Assisted Writing, Move Tag Analysis \u4e00\u3001\u7dd2\u8ad6 \u81ea\u7136\u8a9e\u8a00\u8655\u7406\u662f\u8fd1\u5e7e\u5e74\u5b78\u8853\u6240\u95dc\u5fc3\u7684\u8b70\u984c\uff0c\u5728\u79d1\u6280\u5c1a\u672a\u767c\u5c55\u4ee5\u524d\uff0c\u8a9e\u8a00\u8655\u7406\u5e7e\u4e4e\u9760\u4eba\u529b\u6aa2 \u67e5\u8207\u6821\u6b63\u62fc\u5b57\u8207\u6587\u6cd5\u932f\u8aa4\uff0c\u4f46\u662f\u9760\u4eba\u529b\uff0c\u5247\u6703\u7522\u751f\u4eba\u70ba\u7684\u5931\u8aa4\uff0c\u610f\u601d\u662f\u6307\u4e26\u975e\u4eba\u5de5\u6aa2\u67e5\u5c31\u8868\u793a \u5beb\u4f5c\u7684\u7528\u8a5e\u8207\u8a9e\u6cd5\u6b63\u78ba\uff0c\u6240\u4ee5\u63a1\u7528\u6a5f\u5668\u5b78\u7fd2\u4f86\u66ff\u4ee3\u4eba\u5de5\u7684\u65b9\u5f0f\uff0c\u76f8\u8f03\u65bc\u6a5f\u5668\u5b78\u7fd2\uff0c\u4eba\u5de5\u6821\u6b63\u6216 \u662f\u8655\u7406\u6587\u5b57\u76f8\u5c0d\u82b1\u8cbb\u8f03\u591a\u6642\u9593\u3002 \u82f1\u6587\u662f\u5728\u5b78\u8853\u4e0a\u4e3b\u8981\u6e9d\u901a\u7684\u8a9e\u8a00\uff0c\u6240\u4ee5\u975e\u82f1\u8a9e\u9ad4\u7cfb\u7684\u570b\u5bb6\uff0c\u5c0d\u65bc\u82f1\u6587\u5beb\u4f5c\u9019\u4e00\u90e8\u5206\u76f8\u8f03\u4e4b \u4e0b\uff0c\u767c\u751f\u6587\u6cd5\u8207\u62fc\u5b57\u7684\u932f\u8aa4\u7387\u6703\u660e\u986f\u63d0\u9ad8\uff0c\u56e0\u6b64\u5728\u8cc7\u8a0a\u767c\u9054\u7684\u4e16\u4ee3\uff0c\u5b78\u8853\u6a5f\u69cb\u958b\u59cb\u6536\u96c6\u5beb\u4f5c\u8cc7 \u6599\uff0c\u8b6c\u5982\uff1a\u82f1\u6587\u6aa2\u5b9a\u8003\u7684\u4f5c\u6587 (ETS)\u3001\u5b78\u751f\u5beb\u7684\u4f5c\u6587\u8cc7\u6599\u96c6 (CLEC) \u8207\u7dad\u57fa\u767e\u79d1\u7684\u7de8\u8f2f\u7d00\u9304\u7b49 \u7b49\uff0c\u6709\u9019\u4e9b\u8a9e\u6599\u96c6 (Corpus)\uff0c\u5b78\u8005\u5011\u958b\u59cb\u5f9e\u4e8b\u591a\u65b9\u9762\u7684\u8a9e\u8a00\u8655\u7406\u8207\u5206\u6790\u7814\u7a76\u3002\u5229\u7528\u8a9e\u6599\u5eab\uff0c\u5206 \u6790\u82f1\u8a9e\u7684\u7528\u6cd5 (\u642d\u914d\u8a5e\u3001\u6587\u6cd5)\uff0c\u904b\u7528\u7d71\u8a08\uff0c\u627e\u51fa\u5927\u90e8\u5206\u4eba\u5011\u6240\u4f7f\u7528\u7684\u53e5\u6cd5\uff0c\u5617\u8a66\u8457\u5f9e\u6578\u64da\u7576\u4e2d \u627e\u5230\u7406\u8ad6\uff0c\u85c9\u6b64\u5e6b\u52a9\u5b78\u7fd2\uff0c\u4ee5\u53ca\u63d0\u5347\u5beb\u4f5c\u4e0a\u7684\u6548\u7387\u3002 \u5728\u5b78\u8853\u8ad6\u6587\u4e2d\uff0c\u7c21\u4ecb\u6b64\u4e00\u7ae0\u7bc0\uff0c\u901a\u5e38\u6703\u63cf\u8ff0\uff1a\u554f\u984c\u7684\u80cc\u666f\u3001\u4e3b\u8981\u76ee\u7684\u3001\u89e3\u6c7a\u65b9\u6cd5\u3001\u7d50\u679c\u8207 \u7d50\u8ad6\uff0c\u6b64\u4fee\u8a5e\u7d50\u69cb\u7684\u7d44\u6210\u7a31\u4e4b\u70ba\u300c\u6587\u6b65\u300d \u3002\u5728\u904e\u53bb\u7684\u7814\u7a76\u4e2d [1-3]\uff0c\u91dd\u5c0d\u8ad6\u6587\u7c21\u4ecb\u5b9a\u7fa9\u51fa\u56db\u500b \u6587\u6b65\uff0c\u5305\u62ec\uff1a\u554f\u984c (Problem)\u3001\u65b9\u6cd5 (Solution)\u3001\u8a55\u4f30 (Evaluation) \u8207\u7d50\u8ad6 (Conclusion) \u7b49\u90e8 \u5206\u3002\u7f8e\u570b\u570b\u5bb6\u6a19\u6e96\u5354\u6703 (American National Standard Institute, ANSI) [4]\uff0c\u5be9\u6838\u4e26\u898f\u7bc4\u5beb\u4f5c\u7684 \u6587\u6b65\u7d50\u69cb\u70ba\u76ee\u7684 (Problem)\u3001\u65b9\u6cd5 (Method)\u3001\u7d50\u679c (Result) \u8207\u7d50\u8ad6 (Conclusion)\u3002Swales [5] \u5b9a\u7fa9\u5728\u8ad6\u6587\u5beb\u4f5c\u4f9d\u5faa\u7684\u4e09\u5927\u6587\u6b65\u4fee\u8fad\u7d50\u69cb (Creating a Research Space, CARS)\uff0c\u5305\u62ec\uff1a\u70ba\u5efa \u7acb\u7814\u7a76\u9818\u57df (Establishing a research territory)\u3001\u5efa\u7acb\u5229\u57fa (Establishing a niche)\u3001\u5360\u9818\u5229\u57fa (Occupying the niche)\uff0c\u4e26\u5728\u6bcf\u4e00\u500b\u6587\u6b65\u4fee\u8fad\u7d50\u69cb\u4e4b\u4e0b\u5b9a\u7fa9\u7d30\u7bc0\uff0c\u85c9\u6b64\u5e6b\u52a9\u63cf\u8ff0\u6587\u7ae0\u5167\u5bb9\u3002 \u7279\u5225\u91dd\u5c0d\u5b78\u8853\u82f1\u6587\u5beb\u4f5c\uff0cGlasman-Deal [6] \u63d0\u51fa\u5beb\u4f5c\u4e0a\u6587\u6b65\u6a21\u7d44\uff0c\u5305\u62ec\uff1a\u4ecb\u7d39\u3001\u65b9\u6cd5\u3001\u7d50 \u679c\u3001\u8a0e\u8ad6\u3002Weissberg & Buker [7] \u5b9a\u7fa9\u5b78\u8853\u8ad6\u6587\u5beb\u4f5c\u6587\u6b65\u70ba BPMRC\uff0c\u5373\u80cc\u666f (Background, B)\u3001\u76ee\u7684 (Purpose, P )\u3001\u65b9\u6cd5 (Method, M )\u3001\u7d50\u679c (Result, R)\u3001\u8a0e\u8ad6 (Conclusion, C)\u3002 \u5728\u672c\u7bc7\u8ad6\u6587\u4f7f\u7528 Weissberg & Buker \u63d0\u51fa\u6587\u7ae0\u7684\u6587\u6b65\u67b6\u69cb (\u80cc\u666f\u3001\u76ee\u7684\u3001\u65b9\u6cd5\u3001\u7d50\u679c\u3001 \u7d50\u8ad6)\uff0c\u5229\u7528\u5927\u91cf\u7684\u5b78\u8853\u8ad6\u6587\u8cc7\u6599 (CiteSeerX) \u8207\u5c11\u91cf\u521d\u59cb\u898f\u5247\uff0c\u8a13\u7df4\u8c9d\u6c0f\u6a21\u578b (Bayesian approach)\uff0c\u5b78\u7fd2\u5982\u4f55\u5224\u5225\u53e5\u5b50\u6240\u5c6c\u7684\u6587\u6b65\u3002 \u70ba\u4e86\u5f97\u77e5\u8a13\u7df4\u5b8c\u7562\u7684\u8c9d\u6c0f\u6a21\u578b\u6240\u63d0\u4f9b\u6587\u6b65\u7684\u7cbe\u78ba\u5ea6\uff0c\u5247\u5229\u7528\u55ae\u4e00\u7bc7\u65b0\u7684\u5b78\u8853\u8ad6\u6587\u7c21\u4ecb\uff0c\u900f \u904e\u8c9d\u6c0f\u5206\u985e\u5668\u9032\u884c\u6587\u6b65\u6a19\u5b9a\uff0c\u6700\u7d42\u7531\u4eba\u70ba\u5224\u5225\u6587\u6b65\u7684\u6b63\u78ba\u6027\u3002 \u672c\u8ad6\u6587\u63a5\u8457\u7684\u90e8\u5206\u6703\u5148\u63a2\u8a0e\u76f8\u95dc\u7814\u7a76 (Section 2)\uff0c\u9032\u800c\u63cf\u6558\u5206\u985e\u5668\u81ea\u52d5\u5b78\u7fd2\u6a19\u5b9a\u6587\u6b65\u7684\u904e \u521d\u59cb\u53e5\u6578\u7684\u5206\u5e03\u5982\u8868 4\u3002 5)\u3002 \u6240\u4ee5\u5728\u5be6\u9a57\u7576\u4e2d\uff0c\u9078\u51fa 155 \u500b\u8a5e\u5f59\u7247\u8a9e\u4f5c\u70ba N-\u9023\u8a5e (N-gram)\uff0c\u7576\u4f5c\u521d\u59cb\u7684\u7279\u5fb5\u53c3\u6578\uff0c\u5176 \u7a0b (Section 3)\uff0c\u8207\u5be6\u9a57\u8a2d\u8a08\u3001\u7d50\u679c (Section 4)\u3002\u6700\u5f8c\uff0c\u8a0e\u8ad6\u672a\u4f86\u7684\u7814\u7a76\u65b9\u5411\u8207\u7d50\u8ad6 (Section \u7d50\u679c (Result)\u3001\u8a0e\u8ad6 (Discussion)\uff0c\u8a31\u591a\u5b78\u8005\u4e5f\u5b9a\u7fa9\u51fa\u4e0d\u540c\u7684\u8ad6\u6587\u6587\u6b65\u7d50\u69cb\uff0c\u4f8b\u5982 Swales [5] \u70ba \u7c21\u4ecb\u6b64\u5c0f\u7bc0\u63d0\u51fa CARS(Creating a Research Space) \u6a21\u7d44\uff0cCARS \u4e3b\u8981\u70ba 3 \u5927\u6587\u6b65\u4e26\u7d30\u5206\u70ba 11 \u6587\u6b65\uff0c\u4f7f\u5f97\u8a31\u591a\u5b78\u8005\u4f7f\u7528 CARS \u6a21\u7d44\u63a2\u8a0e\u5beb\u4f5c\u4e0a\u7684\u4fee\u8fad\u65b9\u6cd5\uff0cWeissberg & Buker [7] \u6574\u7406\u51fa BPMRC \u6587\u6b65\u7d50\u69cb\uff0c\u5373\u80cc\u666f (Background)\u3001\u76ee\u7684 (Purpose)\u3001\u65b9\u6cd5 (Method)\u3001\u7d50\u679c (Result)\u3001 \u7d50\u8ad6 (Conclusion)\uff0c\u70ba\u5b78\u8005\u8207\u4f5c\u8005\u63d0\u4f9b\u7814\u7a76\u65b9\u5411\u8207\u5beb\u4f5c\u5efa\u8b70\u3002 \u8fd1\u5e7e\u5e74\u4f86\uff0c\u6709\u8a31\u591a\u5b78\u8005\u63a1\u7528\u4e0d\u540c\u6a5f\u5668\u5b78\u7fd2\u7684\u65b9\u5f0f\u8a13\u7df4\u6587\u6b65\u5206\u985e\u5668\uff0c\u4f8b\u5982 Teufel & Moens [9] \u5229\u7528\u7c21\u6613\u7684\u8c9d\u6c0f\u5206\u985e\u5668 (Naive Bayesian Model, NBM) \u900f\u904e\u4fee\u8fad\u7684\u72c0\u614b\u8207\u95dc\u806f\u91dd\u5c0d\u8ad6\u6587\u5168\u6587 \u9032\u884c\u6587\u6b65\u5206\u985e\u3002Ling [10] \u63d0\u51fa\u96b1\u99ac\u53ef\u592b\u6a21\u578b (Hidden Markov Model, HMM) \u5229\u7528\u7d71\u8a08\u6a5f\u7387\u53bb \u505a\u6587\u6b65\u6a19\u8a3b\uff0cWu & Jason S. [11] \u63d0\u51fa\u4e00\u5957\u7cfb\u7d71 (CARE)\uff0c\u5229\u7528 HMM \u6a19\u8a18\u6587\u6b65\u3002Shimbo [3] \u900f\u904e MEDLINE\uff0c\u63d0\u51fa\u4e00\u5957\u7cfb\u7d71\uff0c\u8b93\u4f7f\u7528\u8005\u53ef\u4ee5\u641c\u5c0b\u7c21\u4ecb\u7279\u5b9a\u7684\u6587\u6b65\uff0c\u6b64\u7cfb\u7d71\u5229\u7528\u652f\u6490\u5411\u91cf\u6a5f \u65b9\u6cd5\u3001\u7d50\u679c) \u3001\u8a0e\u8ad6(\u548c\u524d\u4eba\u7814\u7a76\u7684\u6bd4\u8f03\u8207\u5c0d\u7167)\u8207\u6587\u7bc0\u7d50\u69cb(\u8ad6\u6587\u7d44\u7e54\u3001\u5716\u8868\u7684\u6307\u793a\u3001\u5167\u5bb9\u7684 \u9810\u544a\u8207\u56de\u9867)\u7b49\u56db\u7a2e\u6587\u6b65\uff0c\u800c\u672c\u7bc7\u6240\u63a1\u7528\u7684\u6587\u6b65\u70ba\u4e94\u7a2e (\u80cc\u666f\u3001\u76ee\u7684\u3001\u65b9\u6cd5\u3001\u7d50\u679c\u3001\u7d50\u8ad6)\uff0c\u5728 \u61c9\u7528\u4e0a\uff0c\u8a13\u7df4\u6587\u6b65\u5206\u985e\u5668\u7684\u6f14\u7b97\u6cd5\u6709\u4e9b\u5dee\u5225\uff0c\u672c\u6587\u662f\u63a1\u7528\u8c9d\u6c0f\u5206\u985e (Bayesian) \u800c Guan-Cheng Huang \u63d0\u51fa\u6700\u5927\u71b5\u6a21\u578b (Maximum Entropy, ME)\uff0c\u5dee\u5225\u5728\u65bc\u8c9d\u6c0f\u5728\u904b\u7b97\u7684\u4e00\u958b\u59cb\u9700\u8981\u5148\u9a57\u6a5f \u7387\u689d\u4ef6\uff0c\u4f9d\u64da\u5148\u9a57\u689d\u4ef6\u63a8\u7406\u51fa\u6587\u6b65\u6a5f\u7387\uff0c\u800c\u6700\u5927\u71b5\u6a21\u578b\u5247\u4e0d\u9700\u5148\u9a57\u689d\u4ef6\uff0c\u6240\u4ee5\u6703\u5e73\u5747\u5206\u4f48\uff0c\u4e0d \u50be\u5411\u65bc\u4efb\u4f55\u6587\u6b65\uff0c\u4f46\u5728\u8a13\u7df4\u904e\u7a0b\u4e2d\u63a5\u89f8\u5230\u5176\u4ed6\u8a0a\u606f\uff0c\u5247\u6703\u8abf\u6574\u6587\u6b65\u7684\u6a5f\u7387\u5206\u4f48\u3002 \u76f8\u5c0d\u65bc\u524d\u4eba\u7814\u7a76\u6587\u6b65\u5206\u6790\u7684\u6587\u737b\uff0c\u5728\u672c\u6587\u7576\u4e2d\u63d0\u51fa\u4e00\u5957\u81ea\u52d5\u5b78\u7fd2\u7cfb\u7d71\uff0c\u5229\u7528\u5c08\u5bb6\u5df2\u7d93\u6b78\u7d0d \u7684\u6587\u6b65\u7247\u8a9e\u6574\u7406\u6210 N-\u9023\u8a5e (n-gram)\uff0c\u4ee5\u964d\u4f4e\u4eba\u5de5\u6a19\u793a\u7684\u6210\u672c\uff0c\u5728\u8a13\u7df4\u7684\u904e\u7a0b\u4e2d\uff0c\u5229\u7528\u6587\u6b65\u7279 \u5fb5\uff0c\u81ea\u52d5\u5c07\u53e5\u5b50\u6a19\u793a\uff0c\u4f7f\u5f97\u7cfb\u7d71\u53ef\u4ee5\u5206\u985e\u6587\u6b65\u4e26\u5f9e\u4e2d\u64f4\u5145\u5b57\u8a5e\uff0c\u5229\u7528\u81ea\u52d5\u5316\u6587\u6b65\u6a19\u793a\u800c\u5f97\u5230\u7684 \u5b57\u8a5e\uff0c\u5957\u7528\u5230\u82f1\u6587\u8f14\u52a9\u5beb\u4f5c\u7cfb\u7d71\uff0c\u5e6b\u52a9\u5b78\u751f\u5beb\u5b78\u8853\u8ad6\u6587\u3002 \u4e09\u3001\u65b9\u6cd5 \u70ba\u4e86\u63d0\u4f9b\u4f7f\u7528\u8005\u5728\u5beb\u5b78\u8853\u8ad6\u6587\u6642\uff0c\u5728\u4e0d\u540c\u7ae0\u7bc0 (\u6587\u6b65) \u53ef\u4ee5\u4f7f\u7528\u8f03\u6b63\u78ba\u7684\u5b57\u8a5e\uff0c\u6211\u5011\u5fc5\u9808\u64c1 \u6709\u5927\u91cf\u5df2\u7d93\u88ab\u6a19\u8a3b\u7684\u6587\u6b65\u5b57\u8a5e\u4f86\u505a\u5beb\u4f5c\u4e0a\u7684\u63d0\u793a\uff0c\u800c\u4eba\u5de5\u81ea\u884c\u6a19\u8a3b\u5b57\u8a5e\u7684\u6587\u6b65\u9700\u82b1\u8cbb\u5927\u91cf\u7684\u6642 \u9593\uff0c\u56e0\u6b64\uff0c\u6211\u5011\u63a1\u53d6\u5c08\u5bb6\u6574\u7406\u904e\u7684\u5b57\u8a5e\u900f\u904e\u81ea\u52d5\u5b78\u7fd2\u7684\u65b9\u6cd5\uff0c\u7701\u53bb\u4eba\u5de5\u6a19\u8a3b\u6240\u9700\u82b1\u7684\u6642\u9593\uff0c\u6211 \u5011\u5c07\u554f\u984c\u5b9a\u7fa9\u5982\u4e0b\u3002 \u6211\u5011\u5c07\u53e5\u5b50\u7d93\u904e Genia Tagger \u65b7\u5b57\u4e4b\u5f8c\uff0c\u63a1\u7528\u4e09\u7a2e\u7279\u5fb5\u8a13\u7df4\u51fa\u8c9d\u6c0f\u6a21\u578b (OW, BF, BPC)\uff0c \u4ee5\u8fed\u4ee3 (bootstrapping) \u7684\u65b9\u6cd5\u64f4\u589e\u8c9d\u6c0f\u6a21\u578b\uff0c\u8a08\u7b97\u4e4b\u5f8c\uff0c\u5c07\u4e00\u7bc7\u6587\u7ae0\u7576\u4e2d\u7684\u53e5\u5b50\uff0c\u55ae\u7368\u89c0\u5bdf \u4e00\u7a2e\u6587\u6b65\uff0c\u627e\u51fa\u5728\u6b64\u6587\u6b65\u6a5f\u7387\u6700\u9ad8\u7684\u53e5\u5b50\u9032\u884c\u6587\u6b65\u6a19\u8a3b\uff0c\u907f\u514d\u4e00\u7bc7\u6587\u7ae0\u7576\u4e2d\u53ea\u6709\u4e00\u7a2e\u6587\u6b65\u7684\u60c5 \u4ecb\u7d39\u7cfb\u7d71\u67b6\u69cb\u5716\u8207\u6a21\u7d44\u8a13\u7df4\u7684\u904e\u7a0b\u3002 3.1 \u7cfb\u7d71\u67b6\u69cb \u5716 1: \u7cfb\u7d71\u67b6\u69cb (System architecture) \u5728\u6e2c\u8a66\u968e\u6bb5\uff0c\u9078\u53d6\u65b0\u7684\u4e00\u7bc7\u6587\u7ae0\uff0c\u4e00\u6a23\u4f7f\u7528 Genia Tagger \u9032\u884c\u9810\u5148\u8655\u7406\uff0c\u5c07\u8a13\u7df4\u968e\u6bb5\u5f97\u5230 \u5927\u91cf\u88ab\u6a19\u8a18\u6587\u6b65\u7684 N-\u9023\u8a5e\uff0c\u7576\u4f5c\u5148\u9a57\u8cc7\u8a0a\uff0c\u6e2c\u8a66\u6587\u7ae0\u7d93\u904e\u8a08\u7b97\u4e4b\u5f8c\uff0c\u9010\u53e5\u6240\u5f97\u7684\u6587\u6b65\u6a19\u7c64\u662f \u5426\u6b63\u78ba\uff0c\u9032\u800c\u5f97\u77e5\u65b9\u6cd5\u7684\u6548\u7387\u3002 3.2 \u7279\u5fb5\u9078\u53d6 Original word Base form POS Chunk Named entity (NE) Glyoxysomal Glyoxysomal JJ B-NP B-protein citrate citrate NN I-NP I-protein synthase synthase NN I-NP O in in IN B-PP O pumpkin pumpkin NN B-NP O is be VBZ B-VP O synthesized synthesize VBN I-VP O 3.3 \u521d\u59cb\u898f\u5247 \u91dd\u5c0d\u521d\u59cb N-\u9023\u8a5e\u7684\u9078\u7528\uff0c\u63a1\u7528 Glasman-Deal \u6240\u64b0\u5beb\u7684\u6559\u79d1\u66f8 [6]\uff0c\u6b64\u66f8\u6b78\u7d0d\u51fa\u5728\u4e0d\u540c\u6587\u6b65\u4e0a \u8a72\u5982\u4f55\u5efa\u7acb\u4e00\u500b\u5beb\u4f5c\u67b6\u69cb\u8207\u5728\u6587\u6b65\u4e0a\u6240\u8a72\u4f7f\u7528\u7684\u8a5e\u5f59\uff0c\u5f9e\u4e2d\u6311\u9078\uff0c\u6211\u5011\u5c07\u9078\u53d6\u51fa\u7684 N-\u9023\u8a5e (\u53c3 \u8003\u8868 2\uff0c\u4f9d\u7167\u6587\u6b65\u7d66\u4e88\u521d\u59cb\u503c\uff0c\u6bd4\u5982\"a basic issue for\" \u5728\u66f8\u4e2d\u7684\u5efa\u8b70\u5728\u80cc\u666f (B) \u7576\u4e2d\u4f7f\u7528\uff0c\u6240 \u4ee5\u4ee3\u8868\u6b64 N-\u9023\u8a5e\u5728\u80cc\u666f\u51fa\u73fe\u6b21\u6578\u70ba 1(\u8868 3)\u3002\u5229\u7528\u88ab\u6a19\u8a3b\u5206\u985e\u7684 N-\u9023\u8a5e\uff0c\u7576\u4f5c\u8a13\u7df4\u8cc7\u6599\uff0c\u904b\u7528 \u8c9d\u6c0f\u5b9a\u7406\u8a08\u7b97\u800c\u81ea\u52d5\u7522\u751f\u5927\u91cf\u6a19\u8a3b\u5b8c\u7684\u8ad6\u6587\u53e5\u5b50\uff0c\u5c07\u53e5\u5b50\u5206\u70ba N-\u9023\u8a5e\uff0c\u56de\u994b\u65bc\u521d\u59cb\u503c\uff0c\u7576\u4f5c \u4e0b\u4e00\u6b21\u8a13\u7df4\u8cc7\u6599\uff0c\u800c\u6700\u5f8c\u5c07\u8a13\u7df4\u5b8c\u7684\u7d50\u679c\uff0c\u9032\u4e00\u6b65\u7684\u5206\u6790\u3002 \u2026 in future it is \u8868 3: \u5c07\u521d\u59cb\u898f\u5247\u7d66\u4e88\u51fa\u73fe\u6b21\u6578\uff0c\u7a31\u4e4b\u70ba Count table (CT ) Pattern B P M R C a basic issue for 1 0 0 0 0 approach was developed by 0 1 0 0 0 majority of the tests 0 0 1 0 0 \u2026 \u2026 \u2026 \u2026 \u2026 \u2026 in future it is 0 0 0 0 1 3.4 \u8c9d\u6c0f\u65b9\u6cd5 \u9996\u5148\uff0c\u8c9d\u6c0f\u5b9a\u7406\u9700\u8981\u6709\u5148\u9a57\u6a5f\u7387\uff0c\u624d\u80fd\u8a08\u7b97\u8207\u5206\u6790\uff0c\u6240\u4ee5\u5229\u7528\u66f8\u672c\u5728\u6587\u6b65\u67b6\u69cb\u4e0a\u63a8\u85a6\u7684\u5beb\u6cd5\uff0c \u7576\u4f5c\u8c9d\u6c0f\u7684\u7279\u5fb5\u53c3\u6578\uff0c\u518d\u4f86\uff0c\u5c07\u4e00\u7bc7\u6587\u7ae0\u7684\u7c21\u4ecb\u7576\u4e2d\u7684\u5b57\u8a5e\u7d93\u904e\u8655\u7406\uff0c\u4f7f\u5f97\u6587\u7ae0\u7576\u4e2d\u7684\u7279\u6b8a\u7b26 \u865f\u4e0d\u5f71\u97ff\u55ae\u5b57\uff0c\u5229\u7528\u8c9d\u6c0f\u5b9a\u7406\u8a08\u7b97\u6587\u7ae0\u4e2d\u7684\u6bcf\u500b\u53e5\u5b50\u53bb\u9810\u6e2c\u70ba\u80cc\u666f\u3001\u76ee\u7684\u3001\u65b9\u6cd5\u3001\u7d50\u679c\u3001\u7d50\u8ad6 \u7684\u6a5f\u7387\uff0c\u7576\u4e00\u7bc7\u6587\u7ae0\u7684\u6240\u6709\u53e5\u5b50\u90fd\u8a08\u7b97\u5b8c\u7562\uff0c\u624d\u5f9e\u6240\u6709\u53e5\u5b50\u7576\u4e2d\u662f\u95dc\u65bc\u80cc\u666f\u6b64\u6587\u6b65\u6700\u5927\u503c\u7684\u53e5 \u5b50\u6a19\u8a3b\u70ba\u80cc\u666f\uff0c\u5df2\u88ab\u6a19\u8a3b\u7684\u53e5\u5b50\u5247\u4e0d\u80fd\u91cd\u8907\u88ab\u6a19\u8a3b\uff0c\u7576\u53e5\u5b50\u90fd\u88ab\u6a19\u8a3b\u5b8c\uff0c\u5247\u5c07\u7372\u5f97\u8fa8\u8b58\u7d50\u679c\uff0c \u56de\u994b\u5230\u4e00\u958b\u59cb\u7684\u5148\u9a57\u7279\u5fb5\u53c3\u6578\u3002 \u5f9e Corpus \u53d6\u4e00\u7bc7\u7c21\u4ecb (D 1 )\uff0c\u56e0\u70ba\u6211\u5011\u7684\u5206\u985e\u6a21\u578b\u70ba\u4e94\u500b\u6587\u6b65\uff0c\u82e5\u8a72\u7bc7\u7c21\u4ecb\u6578\u53e5\u5c11\u65bc\u4e94\uff0c \u53e5\u5b50\u6703\u8a08\u7b97\u6bcf\u500b\u6587\u6b65\u7684\u6a5f\u7387\uff0c\u7531\u65bc\u6bcf\u500b\u6587\u6b65\u7684\u8a08\u7b97\u65b9\u6cd5\u90fd\u76f8\u540c\uff0c\u6240\u4ee5\u5728\u5f80\u5f8c\u7684\u6558\u8ff0\u5c07\u5df2\u80cc\u666f (B) \u6587\u6b65\u505a\u4ee3\u8868\u3002 \u800c\u672c\u6587\u4e2d\u7d66\u4e88\u7684\u5148\u9a57\u7279\u5fb5\u53c3\u6578\u662f\u7d66\u4e88 N-\u9023\u8a5e\uff0c\u56e0\u70ba S 1 (2) \u8868 5: S Sentence S 1 ) (4) \u6839\u64da\u4e0a\u8ff0\u7684\u5b9a\u7fa9\uff0c\u5c07\u516c\u5f0f (1)\uff0c\u5b9a\u7fa9\u70ba (5) \u5f62\u767c\u751f\uff0c\u5c07\u5df2\u88ab\u6a19\u5b9a\u6587\u6b65\u7684\u53e5\u5b50\u5206\u70ba N-\u540c\u6642\uff0c\u65b0\u589e\u6216\u66f4\u65b0\u898f\u5247 \u53c3\u8003\u7cfb\u7d71\u67b6\u69cb\u5716 (\u5716 1)\uff0c\u5206\u70ba\u5169\u5927\u90e8\u5206: \u4e00\u90e8\u5206\u662f\u5229\u7528 CiteSeerX \u8a9e\u6599\u5eab\uff0c\u8a13\u7df4\u8c9d\u6c0f\u5206\u985e as as IN B-PP O one one CD B-NP O \u5247\u5ffd\u7565\u8a72\u7bc7\u6587\u7ae0\uff0c\u800c\u53e5\u6578\u8d85\u904e\u4e94\u53e5\uff0c\u5c31\u5b9a\u70ba\u4e00\u7bc7\u5b8c\u6574\u7684\u7c21\u4ecb\uff0c\u800c\u9032\u4e00\u6b65\u5206\u6790\u3002\u5982\u8868 5\u70ba\u64f7\u53d6\u7684 3.4.2 \u8a08\u7b97 N-\u9023\u8a5e\u6a5f\u7387 \u4e00\u7bc7\u5b8c\u6574\u7bc7\u5e45\uff0c\u63a5\u8457\u5c07\u6587\u7ae0\u7576\u4e2d\u7684\u4e00\u500b\u53e5\u5b50 (S 1 ) \u5206\u5225\u8a08\u7b97\u51fa\u53ef\u80fd\u70ba\u80cc\u666f (B)\u3001\u76ee\u7684 (P )\u3001\u65b9\u6cd5 \u5668\uff0c\u53e6\u4e00\u90e8\u5206\u5247\u662f\u4f7f\u7528\u8a13\u7df4\u597d\u7684\u8c9d\u6c0f\u5206\u985e\u5668\u9810\u6e2c\u65b0\u6587\u7ae0\u7684\u53e5\u5b50\u3002 precursor precursor NN I-NP O (M )\u3001\u7d50\u679c (R)\u3001\u7d50\u8ad6 (C) \u7684\u6a5f\u7387\u3002 \u53e5\u5b50\u7d93\u904e\u5206\u5272\u4e4b\u5f8c\uff0c\u5f97\u5230\u5176\u4e2d\u4e00\u6bb5 N-\u9023\u8a5e (ng 1 )\uff0c\u4f8b\u5982\u70ba\"ng 1 : glyoxysomal citrate synthase \u5728\u8a13\u7df4\u968e\u6bb5\uff0c\u5f9e Glasman-Deal \u6b64\u66f8\u7576\u4e2d\u4f9d\u7167\u6587\u6b65\u6240\u63d0\u4f9b\u7684\u8cc7\u8a0a\uff0c\u64f7\u53d6 155 \u53e5 N-\u9023\u8a5e\u7576\u4f5c that that WDT B-NP O in\"\uff0c\u5148\u8a08\u7b97 ng 1 \u51fa\u73fe\u7684\u6a5f\u7387\u3002 \u521d\u59cb\u6587\u6b65\u8a13\u7df4\u898f\u5247\uff0c\u800c\u5728\u8cc7\u6599\u65b9\u9762\u5247\u662f\u53d6\u5c08\u9580\u6536\u96c6\u5b78\u8853\u8ad6\u6587\u7c21\u4ecb\u7684\u8a9e\u6599\u5eab (CiteSeerX)\uff0c\u9808\u5148 \u5c07\u8a9e\u6599\u5eab\u9010\u7bc7\u7d93\u904e Genia Tagger \u9032\u884c\u65b7\u5b57\u8655\u7406\uff0c\u5229\u7528\u7a0b\u5f0f\u5c07\u7c21\u4ecb\u4e2d\u7684\u53e5\u5b50\u5206\u5272\u6210\u4e00\u53e5\u4e00\u53e5\uff0c\u7136 \u5f8c\u518d\u628a\u53e5\u5b50\u4f9d\u7167 Genia Tagger \u63d0\u4f9b\u7684\u8a5e\u6027\u6a19\u8a3b (Part of Speech, POS)\u3001\u5b57\u6839\u9084\u539f (Base form) \u904e\u5206\u6790\u5c07\u8a9e\u6599\u5eab\u6240\u63d0\u4f9b\u53e5\u5b50\u5b57\u8a5e\u9032\u884c\u6587\u6b65\u6a19\u8a3b\uff0c\u4e26\u56de\u994b\u5230\u521d\u59cb\u8868\u7576\u4e2d\uff0c\u7d93\u904e\u53cd\u8986\u8a13\u7df4\u7684\u904e\u7a0b\uff0c \u64f4\u589e\u5df2\u88ab\u6a19\u8a18\u7684 N-\u9023\u8a5e\u7576\u4f5c\u4e0b\u4e00\u6b21\u8a08\u7b97\u7684\u4f9d\u64da\u3002 P (move-tag|S 1 ) = P (S 1 ) , when move-tag \u2208 {B, P, M, R, C} (1) P (move-tag) \u00d7 P (S 1 |move-tag) \u8207\u8a9e\u610f\u5340\u584a (chunk) \u505a\u9810\u5148\u8655\u7406\uff0c\u5c07\u8655\u7406\u904e\u5f8c\u7684\u53e5\u5b50\u5206\u70ba N-\u9023\u8a5e\uff0c\u4f9d\u7167\u521d\u59cb\u6587\u6b65\u7576\u4f5c\u4f9d\u64da\uff0c\u7d93 has have VBZ B-VP O \u2026 \u2026 \u2026 \u2026 3.4.1 \u8a08\u7b97\u6587\u6b65\u6a5f\u7387 n-gram B P M R C \u2026 . (Period) . (Period) . O O \u4ee5 S 1 \u70ba\u4f8b\uff0c\u6211\u5011\u5c07\u5206\u5225\u8a08\u7b97\u6587\u6b65\u7684\u6a5f\u7387\u503c\u3002 ng 1 glyoxysomal citrate synthase in</td></tr></table>", |
| "num": null, |
| "type_str": "table" |
| }, |
| "TABREF1": { |
| "text": "\u6211\u5011\u63a1\u53d6\u4e09\u7a2e\u7279\u5fb5\u8a13\u7df4\u5f97\u5230\u7684\u7d50\u679c\uff0c\u7531\u65bc\u539f\u59cb\u8cc7\u6599 (OW) \u8207\u5b57\u6839\u9084\u539f (BF) \u6b64\u5169\u7a2e\u7279\u5fb5\u6240 \u5f97\u5230\u7684\u7cbe\u78ba\u7387\uff0c\u6c92\u6709\u9810\u671f\u7684\u5dee\u8ddd\uff0c\u800c\u5728\u8a5e\u6027\u66ff\u63db (BPC) \u7684\u7279\u5fb5\u4e0b\uff0c\u76f8\u5c0d\u65bc OW \u8207 BF\uff0c\u6587\u6b65 \u7684\u7cbe\u78ba\u7387\u6709\u660e\u986f\u63d0\u5347\uff0c\u53ef\u80fd\u56e0\u70ba\u63d0\u4f9b\u7684\u898f\u5247\u8868\u9054\u65b9\u5f0f\u6bd4\u8f03\u7c21\u6613\uff0c\u5728\u63d0\u4f9b\u8a08\u7b97\u6642\u7684\u6587\u6b65\u7279\u5fb5\u8f03\u660e \u986f\u3002 \u8868 12: \u5229\u7528\u539f\u59cb\u8cc7\u6599 (OW) \u6240\u5f97\u7684\u7cbe\u78ba\u7387 (Accuracy) Accuracy 39.43% 41.54% 42.95% 42.95% 43.66% 45.77% \u8868 13: \u5229\u7528\u5b57\u6839\u9084\u539f (BF) \u6240\u5f97\u7684\u7cbe\u78ba\u7387 (Accuracy) Accuracy 40.14% 42.25% 44.36% 45.07% 45.77% \u8868 14: \u5229\u7528\u8a5e\u6027\u6a19\u8a18\u8207\u610f\u5143\u96c6\u7d44\u8655\u7406\u6587\u7ae0 (BPC) \u6240\u5f97\u7684\u7cbe\u78ba\u7387 (Accuracy)", |
| "html": null, |
| "content": "<table><tr><td>4.1 \u8a9e\u6599\u5eab \u7d44\u5408\u904e\u591a\u3002</td><td/><td/><td/></tr><tr><td colspan=\"5\">\u672c\u6587\u91dd\u5c0d\u8f14\u52a9\u82f1\u6587\u5b78\u8853\u8ad6\u6587\u5beb\u4f5c\uff0c\u56e0\u6b64\u6211\u5011\u63a1\u7528\u5c08\u9580\u6536\u96c6\u767c\u8868\u904e\u7684\u5b78\u8853\u8ad6\u6587\u8a9e\u6599\u96c6 (Cite-</td></tr><tr><td colspan=\"5\">SeerX) 1 \u3002CiteSeerX \u662f\u4e00\u500b\u95dc\u65bc\u6587\u737b\u7684\u641c\u5c0b\u5f15\u64ce\uff0c\u5728 1997 \u5e74\uff0c\u7531\u7f8e\u570b\u666e\u6797\u65af\u9813\u5927\u5b78\u958b\u767c CiteSeer\uff0c\u5efa\u7acb\u4e00\u500b\u6578\u4f4d\u5716\u66f8\u9928\uff0c\u7531\u65bc CiteSeer \u53ea\u80fd\u6536\u96c6\u516c\u958b\u7684\u6587\u4ef6\uff0c\u4f7f\u5f97\u6240\u6536\u96c6\u6587\u7ae0\u9818\u57df\u6709 \u8868 11: \u5c08\u5bb6\u6a19\u8a3b\u6587\u6b65\u7684\u53e5\u6578 (# of sentence with correct tags) \u7bc7\u6578 1,000 2,000 3,000 4,000 5,000</td></tr><tr><td colspan=\"5\">\u9650\uff0c\u70ba\u4e86\u514b\u670d\u4fb7\u9650\u6027\uff0c\u91dd\u5c0d\u7cfb\u7d71\u67b6\u69cb\u91cd\u65b0\u5b9a\u5411 (CiteSeerX)\uff0c\u65bc 2007 \u5e74\u63a1\u7528\u6a5f\u5668\u5b78\u7fd2\u7684\u65b9\u6cd5\uff0c move-tag B P M R C Total Sentence</td></tr><tr><td colspan=\"5\">\u81ea\u52d5\u8fa8\u8b58\u7db2\u8def\u4e0a\u5b58\u5728\u7684\u8ad6\u6587\uff0c\u7136\u5f8c\u4f9d\u7167\u7d22\u5f15\u6a19\u793a\u6587\u7ae0\uff0c\u900f\u904e\u5f15\u6587\u7684\u5f71\u97ff\uff0c\u9023\u63a5\u6bcf\u7bc7\u6587\u7ae0\u3002 CiteSeerX \u7e3d\u5171\u64c1\u6709 138 \u842c\u591a\u7bc7\u7684\u6587\u737b\uff0c\u4e3b\u8981\u7684\u5167\u5bb9\u70ba\u79d1\u5b78\u9818\u57df(\u5305\u542b\u8cc7\u5de5\u548c\u751f\u91ab\u9818\u57df) \uff0c # of sentence 27 10 21 60 24 142</td></tr><tr><td colspan=\"5\">\u800c\u9019\u4e9b\u8cc7\u6599\u4f86\u6e90\u901a\u5e38\u70ba PDF \u683c\u5f0f\uff0c\u7d93\u904e\u81ea\u52d5\u8fa8\u8b58\u8f49\u6a94\u6210\u6587\u5b57\uff0c\u56e0\u6b64\u8a9e\u6599\u5eab\u88cf\u982d\u5305\u542b\u8a31\u591a\u63db\u884c</td></tr><tr><td colspan=\"5\">\u9023\u5b57\u7b26\u865f\u3001\u7279\u6b8a\u7b26\u865f\u7b49\u96dc\u8a0a\uff0c\u6240\u4ee5\u5728\u4f7f\u7528\u8cc7\u6599\u4e4b\u524d\uff0c\u6211\u5011\u900f\u904e\u6587\u5b57\u8655\u7406\uff0c\u5c07\u5197\u9918\u7684\u7b26\u865f\u6216\u662f\u65e5 \u671f\u683c\u5f0f\u6368\u53bb\uff0c\u9032\u800c\u5f97\u5230\u8f03\u5b8c\u5584\u7684\u4e00\u7bc7\u8ad6\u6587\u3002 \u5728\u6e2c\u8a66\u968e\u6bb5\uff0c\u70ba\u80fd\u4e86\u89e3\u8cc7\u6599\u7d93\u904e\u8a13\u7df4\u7684\u7bc7\u6578\u662f\u5426\u5f71\u97ff\u6587\u6b65\u6a19\u7c64\u7684\u7cbe\u78ba\u7387\uff0c\u6240\u4ee5\u8a55\u4f30\u6bcf\u7d93\u904e \u4e00\u5343\u7bc7\u8a13\u7df4\u7684 CT \u8868\uff0c\u9810\u6e2c\u53e5\u5b50\u6587\u6b65\u7684\u6a19\u8a3b\u662f\u5426\u6b63\u78ba\u3002 \u7bc7\u6578 1,000 2,000 3,000 4,000 5,000 20,000</td></tr><tr><td colspan=\"5\">4.2 \u5be6\u9a57\u8a2d\u5b9a \u9996\u5148\u8a55\u4f30\u539f\u59cb\u8cc7\u6599\u7d93\u904e\u9010\u7bc7\u8a13\u7df4\u800c\u5f97\u5230\u7684 CT \u8cc7\u6599\u8868\uff0c\u6240\u9810\u6e2c\u53e5\u5b50\u6587\u6b65\u7684\u7cbe\u78ba\u7387\u3002 \u7531\u5716 3\u5f97\u77e5\u8a55\u4f30\u7684\u7d50\u679c\uff0c\u53ef\u4ee5\u89c0\u5bdf\u53e5\u5b50\u6587\u6b65\u6a19\u7c64\u7684\u7cbe\u78ba\u7387\uff0c\u767c\u73fe CT \u8cc7\u6599\u8868\u6bcf\u7d93\u904e\u4e00\u5343\u7bc7\u7684 Accuracy 43.66% 46.47% 49.29% 47.18% 47.88% 56.33%</td></tr><tr><td colspan=\"5\">\u53c3\u8003\u7cfb\u7d71\u67b6\u69cb\u5716 (\u5716 1)\uff0c\u6211\u5011\u5229\u7528\u521d\u59cb\u898f\u5247 (Initial pattern)\uff0c\u5206\u6790\u8a9e\u6599\u5eab (CiteSeerX) \u63d0\u4f9b\u7684 \u8a13\u7df4\uff0c\u5f97\u5230\u7684\u7d50\u679c\u9010\u6f38\u6539\u5584\u3002</td></tr><tr><td colspan=\"5\">\u6587\u7ae0\uff0c\u9010\u7bc7\u8a13\u7df4\u8a9e\u8a00\u6a21\u7d44\uff0c\u6bcf\u7576\u7d93\u904e\u4e00\u5343\u7bc7\u8a13\u7df4\u7684\u8a9e\u8a00\u6a21\u7d44\uff0c\u5247\u6703\u6e2c\u8a66\u7cbe\u78ba\u5ea6\uff0c\u5728\u672c\u8ad6\u6587\u7576 \u7531\u65bc BF \u505a\u51fa\u7684\u5be6\u9a57\u7d50\u679c\u8207 OW \u76f8\u4f3c\uff0c\u6240\u4ee5\u5728\u6b64\u53ea\u986f\u73fe\u7cbe\u78ba\u7387\u7684\u7d50\u679c\u3002</td></tr><tr><td colspan=\"5\">\u4e2d\uff0c\u53d6\u5169\u842c\u7bc7\u7576\u8a13\u7df4\u8cc7\u6599\u3002 \u518d\u8005\uff0c\u8a55\u4f30\u6587\u7ae0\u7d93\u904e\u8a5e\u6027\u6a19\u8a18\u8207\u610f\u5143\u96c6\u7d44\u8655\u7406\u7684\u53e5\u5b50\u6240\u8a13\u7df4\u7684 CT \u8cc7\u6599\u8868\uff0c\u9810\u6e2c\u53e5\u5b50\u6587\u6b65 \u4e94\u3001\u7d50\u8ad6</td></tr><tr><td colspan=\"5\">\u5728\u6e2c\u8a66\u968e\u6bb5\uff0c\u4e8b\u5148\u5f9e\u8a9e\u6599\u5eab\u96a8\u6a5f\u63d0\u51fa 20 \u7bc7\u5c1a\u672a\u7d93\u904e\u8a13\u7df4\u7684\u6587\u7ae0\uff0c\u7d93\u904e\u5c08\u5bb6\u9010\u53e5\u6a19\u8a3b\u6587\u6b65\uff0c \u7684\u7cbe\u78ba\u7387\u3002 \u6211\u5011\u8a2d\u8a08\u4e00\u5957\u8a9e\u8a00\u8a13\u7df4\u65b9\u6cd5\uff0c\u5c08\u9580\u8655\u7406\u5b78\u8853\u8ad6\u6587\uff0c\u9010\u7bc7\u9010\u53e5\u6a19\u4e0a\u9069\u5408\u7684\u6587\u6b65\u6a19\u7c64\uff0c\u5c07\u6536\u96c6</td></tr><tr><td colspan=\"5\">\u6211\u5011\u900f\u904e\u56db\u500b\u5c08\u5bb6\u91dd\u5c0d\u6b64 20 \u7bc7 (\u5171 185 \u53e5) \u9010\u53e5\u7d66\u4e88\u6587\u6b65\u6a19\u7c64\uff0c\u6311\u51fa\u5176\u4e2d\u4e09\u500b\u4eba\u4ee5\u4e0a\u7d66\u4e88\u53e5\u5b50 \u5230\u7684 N-\u9023\u8a5e\u9032\u800c\u6574\u7406\uff0c\u4ee5\u5e6b\u52a9\u5b78\u751f\u5beb\u4f5c\u5b78\u8853\u8ad6\u6587\u3002\u6211\u5011\u6240\u4f7f\u7528\u7684\u65b9\u6cd5\uff0c\u662f\u5229\u7528\u5c08\u5bb6\u63d0\u4f9b\u5728\u7279</td></tr><tr><td colspan=\"5\">\u7684\u6a19\u7c64\u76f8\u540c\u4f86\u8a55\u4f30\u8cc7\u6599\u7684\u6e96\u78ba\u6027 (\u4e09\u4eba\u4ee5\u4e0a\u76f8\u540c\u53e5\u6578\u5171 142 \u53e5)\u3002 \u5b9a\u6587\u6b65\u5e38\u4f7f\u7528\u7684\u5b57\u8a5e\uff0c\u85c9\u4ee5\u900f\u904e\u8a9e\u6599\u5eab\u9032\u884c\u5206\u6790\u4e26\u64f7\u53d6\u53e5\u5b50\u7684\u7279\u5fb5\uff0c\u7522\u751f\u5927\u91cf\u5df2\u6a19\u8a3b\u7684 N-\u9023</td></tr><tr><td colspan=\"5\">\u6211\u5011\u5c07 20 \u7bc7\u6587\u7ae0\u9032\u884c\u6e2c\u8a66\uff0c\u5c07\u53e5\u5b50\u6a19\u4e0a\u6a19\u7c64\u3002\u4e4b\u5f8c\u5b9a\u7fa9\u5982\u4e0b\u7cbe\u78ba\u7387\uff0c\u4f9d\u7167 142 \u53e5\u6b63\u78ba\u7b54 \u8a5e\uff0c\u5c07\u5f97\u5230\u7684\u8a13\u7df4\u8cc7\u6599\uff0c\u61c9\u7528\u5230\u6587\u6b65\u5206\u985e\u5668\u3002</td></tr><tr><td colspan=\"5\">\u6848\uff0c\u627e\u51fa\u6a19\u4e0a\u6b63\u78ba\u6587\u6b65\u7684\u53e5\u6578\u3002 \u5728\u672a\u4f86\u7814\u7a76\u4e2d\uff0c\u6211\u5011\u5c07\u64f7\u53d6\u8fa8\u8b58\u5ea6\u9ad8\u7684\u6587\u6b65\u7279\u5fb5\uff0c\u63d0\u5347\u6587\u6b65\u8fa8\u8b58\u7684\u6e96\u78ba\u7387\u3002\u4f8b\u5982\u8a08\u7b97 N-\u9023</td></tr><tr><td colspan=\"5\">\u8868 9: \u65b0\u589e\u898f\u5247 \u8a5e\u4e4b\u9593\u7684\u76f8\u4f3c\u5ea6\uff0c\u627e\u51fa\u5c6c\u65bc\u6587\u6b65\u7684\u53e5\u578b\uff0c\u627e\u51fa\u7279\u6b8a\u55ae\u5b57\u51fa\u73fe\u7684\u983b\u7387\uff0c\u589e\u5f37\u6587\u6b65\u5c6c\u6027\u7684\u7279\u5fb5\uff0c\u5e0c Accuracy = # of sentences with correct move-tag . (11) \u671b\u80fd\u5728\u5b78\u8853\u8ad6\u6587\u5beb\u4f5c\u4e0a\u63d0\u4f9b\u66f4\u597d\u7684\u5e6b\u52a9\u3002\u540c\u6642\uff0c\u4e26\u8003\u91cf\u6587\u6b65\u7684\u9806\u5e8f\u8207\u4f4d\u7f6e\uff0c\u4f86\u8abf\u6574\u8c9d\u6c0f\u898f\u5247\u3002 142 \u5728\u5be6\u9a57\u65b9\u9762\uff0c\u8003\u616e N-\u9023\u8a5e\u4e2d\uff0c\u4e0d\u540c\u7684 N \u503c\uff0c\u4e5f\u5c07\u7528\u66f4\u591a\u7684\u8a13\u7df4\u8cc7\u6599\uff0c\u4e26\u8207\u5176\u4ed6\u7684\u5206\u985e\u65b9\u6cd5</td></tr><tr><td>4.3 \u5be6\u9a57\u7d50\u679c (\u4f8b\u5982\uff1aSVM, ME)\u6bd4\u8f03\u3002</td><td/><td/><td/></tr><tr><td colspan=\"5\">pattern (4-gram) glyoxysomal citrate synthase in 1 0 B P M R C 0 \u672c\u6587\u5229\u7528 CiteSeerX \u63d0\u4f9b\u7684\u8cc7\u6599\uff0c\u8a08\u7b97\u6bcf\u7d93\u904e\u4e00\u5343\u7bc7\u7684\u8a13\u7df4\u5f8c\uff0c\u5247\u6703\u589e\u52a0\u591a\u5c11 N-\u9023\u8a5e\u7684\u5148\u9a57 0 0 \u2026 \u2026 \u2026 \u2026 \u2026 \u2026 \u898f\u5247\uff0c\u56e0\u8cc7\u6599\u7d93\u904e Genia Tagger \u8655\u7406\u4e4b\u5f8c\u6703\u63d0\u4f9b\u8cc7\u6599\u539f\u59cb\u5b57\u8a5e (Original Word)\u3001\u5b57\u6839\u9084\u539f (Based form)\u3001\u8a5e\u6027\u6a19\u8a18 (POS) \u7b49\u8cc7\u8a0a\uff0c\u5247\u8a13\u7df4\u65b9\u6cd5\u7d66\u7684\u6587\u5b57\u8cc7\u6599\u70ba\u6b64\u4e09\u7a2e\u65b9\u5f0f\uff0c\u900f\u904e\u904b\u7b97\u5f97 \u5716 3: \u95dc\u65bc OW \u6a19\u5b9a\u53e5\u5b50\u6587\u6b65\u8cc7\u8a0a \u5716 4: \u95dc\u65bc BPC \u6a19\u5b9a\u53e5\u5b50\u6587\u6b65\u8cc7\u8a0a</td></tr><tr><td colspan=\"5\">\u5230\u7684\u7d50\u679c\u3002 \u7531\u5716 4\u5f97\u77e5\u8a55\u4f30\u7d50\u679c\uff0c\u76f8\u5c0d\u65bc\u539f\u59cb\u8cc7\u6599\u7684\u7cbe\u78ba\u7387\u7565\u70ba\u63d0\u9ad8\uff0c\u539f\u56e0\u70ba\u5c07\u53e5\u5b50\u7c21\u55ae\u5316\uff0c\u907f\u514d\u610f\u601d</td></tr><tr><td colspan=\"5\">\u82e5 ng \u5b58\u5728\u65bc CT \u4e2d\uff0c\u5247\u6703\u4f9d\u7167 S 1 \u88ab\u6a19\u5b9a\u7684\u7d50\u679c\uff0c\u66f4\u65b0 ng \u5728\u8a72\u6587\u6b65\u51fa\u73fe\u7684\u6b21\u6578 (\u8868 10)\u3002 \u76f8\u540c\u7684 N-\u9023\u8a5e\uff0c\u56e0\u70ba\u4e00\u4e9b\u6578\u5b57\u6216\u662f\u975e\u82f1\u8a9e\u55ae\u5b57\u800c\u964d\u4f4e\u6587\u6b65\u7684\u7279\u5fb5\u8a08\u7b97\u3002</td></tr><tr><td>4.4 \u8a0e\u8ad6</td><td/><td/><td/></tr><tr><td colspan=\"5\">\u8acb\u6ce8\u610f\uff0c\u5728\u9019\u6b65\u9a5f\u4e2d\uff0c\u6211\u5011\u50c5\u662f\u5c07\u524d\u4e00\u6b65\u9a5f\u4e2d\u3001\u7528\u8c9d\u6c0f\u65b9\u6cd5\u5224\u5b9a\u7684\u6587\u6b65\u7d50\u679c(\u6c92\u6709\u4eba\u70ba\u4ecb\u5165 \u5224\u5b9a) \uff0c\u52a0\u56de CT \u4e2d\u3002\u6211\u5011\u4e26\u4e0d\u7acb\u5373\u5224\u5b9a\u6240\u6a19\u5b9a\u7684\u6587\u6b65\u662f\u5426\u70ba\u6b63\u78ba\uff0c\u800c\u662f\u4ee5\u4e0d\u65b7\u8fed\u4ee3(iterative) \u6574\u9ad4\u800c\u8a00\uff0c\u6587\u6b65\u9810\u6e2c\u7684\u6b63\u78ba\u7387\u70ba 56%\uff0c\u5c1a\u6709\u9032\u6b65\u7684\u7a7a\u9593\u3002\u5728\u8a13\u7df4\u898f\u5247\u5c11\u7684\u60c5\u6cc1\u4e0b (155 \u500b\u898f</td></tr><tr><td colspan=\"5\">\u7684\u65b9\u5f0f\uff0c\u5229\u7528\u8c9d\u6c0f\u65b9\u6cd5\uff0c\u4f86\u6293\u4f4f\u8a13\u7df4\u8cc7\u6599(training data)\u7684\u7279\u6027\u3002 \u5247)\uff0c\u80fd\u9054\u5230\u4e00\u534a\u7684\u6e96\u78ba\u7387\uff0c\u5c0d\u65bc\u6b64\u60c5\u5f62\uff0c\u6211\u5011\u4fdd\u6301\u6a02\u89c0\u7684\u614b\u5ea6\u3002</td></tr><tr><td colspan=\"5\">\u800c\u5728\u7d71\u8a08\u898f\u5247\u65b0\u589e\u5716\u8868\u4e2d\uff0c\u5c0d\u65bc\u898f\u5247\u6578\u6301\u7e8c\u589e\u52a0\u7684\u554f\u984c\uff0c\u6211\u5011\u8a2d\u5b9a\u4e09\u7a2e\u7279\u5fb5\u9078\u53d6\u5b57\u8a5e\uff0c\u5c07</td></tr><tr><td colspan=\"5\">\u53e5\u5b50\u7684\u5b57\u8a5e\u8b8a\u5f97\u8f03\u62bd\u8c61\uff0c\u4f8b\u5982\u4e0d\u5728\u4e4e\u6642\u614b\u8207\u8907\u6578\u3001\u66ff\u63db\u7b26\u865f\u6216\u6578\u5b57\u7b49\u7b49\uff0c\u60f3\u85c9\u6b64\u5c07\u898f\u5247\u7684\u6578\u91cf \u8868 10: \u66f4\u65b0\u898f\u5247 \u6e1b\u5c11\uff0c\u4f46\u4e26\u672a\u9054\u5230\u9810\u671f\u7684\u6548\u679c (\u8868 15)\u3002</td></tr><tr><td colspan=\"2\">pattern (4-gram)</td><td>B</td><td colspan=\"2\">P M R C</td></tr><tr><td colspan=\"4\">glyoxysomal citrate synthase in B i + 1 0</td><td>0</td><td>0 0</td></tr><tr><td/><td>\u2026</td><td>\u2026</td><td colspan=\"2\">\u2026 \u2026 \u2026</td></tr><tr><td>\u56db\u3001\u5be6\u9a57</td><td colspan=\"3\">\u5716 2: \u7d93\u7531\u8a13\u7df4\uff0c\u589e\u52a0\u7684\u898f\u5247\u6578</td></tr><tr><td colspan=\"2\">\u5728\u672c\u7bc0\u4e2d\uff0c\u6211\u5011\u5c07\u8a0e\u8ad6\u5be6\u9a57\u8a2d\u5b9a\u8207\u7d50\u679c\u8a0e\u8ad6\u3002 \u7bc7\u6578 1,000 2,000 3,000</td><td>4,000</td><td/><td>5,000</td><td>20,000</td></tr><tr><td colspan=\"5\">\u6bcf\u7d93\u904e\u4e00\u5343\u7bc7\u7684\u8a13\u7df4\uff0c\u5247 CT \u6703\u589e\u52a0\u7d04\u516d\u842c\u7684 N-\u9023\u8a5e\u898f\u5247\uff0c\u7d93\u904e\u6e2c\u8a66\uff0c\u4e26\u6c92\u6709\u767c\u73fe\u6536\u6582\u7684</td></tr><tr><td colspan=\"5\">\u73fe\u8c61\uff0c\u53ef\u80fd\u662f\u56e0\u70ba\u6587\u7ae0\u7576\u4e2d\u6709\u904e\u591a\u7279\u6b8a\u5b57\u8a5e\uff0c\u6216\u8005\u662f\u56e0\u70ba\u6211\u5011\u8a2d\u5b9a\u7684 N-\u9023\u8a5e\u592a\u904e\u65bc\u9577\uff0c\u5c0e\u81f4</td></tr></table>", |
| "num": null, |
| "type_str": "table" |
| } |
| } |
| } |
| } |
