{"id":3587,"date":"2020-09-30T14:52:00","date_gmt":"2020-09-30T06:52:00","guid":{"rendered":"http:\/\/www.sniper97.cn\/?p=3587"},"modified":"2020-09-30T14:52:00","modified_gmt":"2020-09-30T06:52:00","slug":"%e3%80%90%e8%87%aa%e7%84%b6%e8%af%ad%e8%a8%80%e5%a4%84%e7%90%86%e3%80%91%e5%91%bd%e5%90%8d%e5%ae%9e%e4%bd%93%e8%af%86%e5%88%ab","status":"publish","type":"post","link":"http:\/\/www.sniper97.cn\/index.php\/note\/deep-learning\/nlp-base\/3587\/","title":{"rendered":"\u3010\u81ea\u7136\u8bed\u8a00\u5904\u7406\u3011\u547d\u540d\u5b9e\u4f53\u8bc6\u522b"},"content":{"rendered":"\n<h2 class=\"wp-block-heading\">\n\u547d\u540d\u5b9e\u4f53\u8bc6\u522b\n<\/h2>\n\n\n<p>\u547d\u540d\u5b9e\u4f53\u8bc6\u522b\uff08Named Entity Recognition\uff0cNER\uff09\uff0c\u5b83\u662f\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4e2d\u7684\u4e00\u9879<strong>\u57fa\u7840\u4efb\u52a1<\/strong>\uff0c\u5e94\u7528\u8303\u56f4\u975e\u5e38\u5e7f\u6cdb\u3002<\/p>\n\n\n<p>\u547d\u540d\u5b9e\u4f53\u4e00\u822c\u662f\u6587\u4e2d\u5177\u6709\u7279\u5b9a\u610f\u4e49\u6216\u8005\u6307\u4ee3\u6027\u5f3a\u7684\u5b9e\u4f53\uff0c\u901a\u5e38\u5305\u62ec\uff1a\u4eba\u540d\u3001\u5730\u540d\u3001\u7ec4\u7ec7\u673a\u6784\u540d\u3001\u65e5\u671f\u65f6\u95f4\u3001\u4e13\u6709\u540d\u8bcd\u7b49\u3002<\/p>\n\n\n<p>NER\u7cfb\u7edf\u5c31\u662f\u4ece\u7ed3\u6784\u5316\u7684\u8f93\u5165\u6587\u672c\u4e2d\u62bd\u53d6\u51fa\u4e0a\u8ff0\u5b9e\u4f53\uff0c\u5e76\u4e14\u53ef\u4ee5\u6309\u7167\u4e1a\u52a1\u9700\u6c42\u8bc6\u522b\u51fa\u66f4\u591a\u7c7b\u522b\u7684\u5b9e\u4f53\uff0c\u6bd4\u5982\u4ea7\u54c1\u540d\u79f0\u3001\u578b\u53f7\u3001\u4ef7\u683c\u7b49\u3002<\/p>\n\n\n<p>\u5728\u5b66\u672f\u4e0aNER\u6240\u6d89\u53ca\u7684\u547d\u540d\u5b9e\u4f53\u4e00\u822c\u5305\u62ec\u4e09\u5927\u7c7b\uff08\u5b9e\u4f53\u7c7b\u3001\u65f6\u95f4\u7c7b\u3001\u6570\u5b57\u7c7b\uff09\u548c\u4e03\u5c0f\u7c7b\uff08\u4eba\u540d\u3001\u5730\u540d\u3001\u7ec4\u7ec7\u673a\u6784\u540d\u3001\u65f6\u95f4\u3001\u65e5\u671f\u3001\u8d27\u5e01\u3001\u767e\u5206\u6bd4\uff09\u3002<\/p>\n\n\n<p>\u540c\u65f6NER\u4e5f\u662f\u5173\u7cfb\u62bd\u53d6\u3001\u65f6\u95f4\u62bd\u53d6\u3001\u77e5\u8bc6\u56fe\u8c31\u3001\u673a\u5668\u7ffb\u8bd1\u3001\u95ee\u7b54\u7cfb\u7edf\u7b49\u8bf8\u591aNLP\u4efb\u52a1\u7684\u57fa\u7840\u3002<\/p>\n\n\n<h2 class=\"wp-block-heading\">\u4e2d\u6587\u547d\u540d\u5b9e\u4f53\u8bc6\u522b\u7684\u6311\u6218<\/h2>\n\n\n<p>1.\u6ca1\u6709\u660e\u786e\u7684\u754c\u9650\uff08\u82f1\u6587\u5355\u8bcd\u4ee5\u7a7a\u683c\u4e3a\u754c\u9650\uff09\u3002<\/p>\n\n\n<p>2.\u4e0d\u50cf\u82f1\u6587\u6709\u5927\u5c0f\u5199\u7684\u5f62\u6001\u6307\u793a\uff08\u6bd4\u5982\u4e13\u5c5e\u540d\u8bcd\u9996\u5b57\u6bcd\u5927\u5199\uff09\u3002<\/p>\n\n\n<p>3.\u4e2d\u6587\u7528\u5b57\u7075\u6d3b\u591a\u53d8\uff0c\u6709\u4e9b\u8bcd\u5728\u8131\u79bb\u4e0a\u4e0b\u6587\u8bed\u5883\u60c5\u51b5\u4e0b\u65e0\u6cd5\u5224\u65ad\u662f\u5426\u662f\u547d\u540d\u5b9e\u4f53\u4e14\u5728\u4e0d\u540c\u7684\u4e0a\u4e0b\u6587\u8bed\u5883\u4e2d\u53ef\u80fd\u662f\u4e0d\u540c\u7684\u5b9e\u4f53\u7c7b\u578b\u3002<\/p>\n\n\n<p>4.\u547d\u540d\u5b9e\u4f53\u5b58\u5728\u5d4c\u5957\u73b0\u8c61\u3002\u5982\u2018\u5927\u8fde\u7406\u5de5\u5927\u5b66\u8f6f\u4ef6\u5b66\u9662\u2019\uff0c\u5728\u8be5\u5b9e\u4f53\u4e2d\u5d4c\u5957\u4e86\u5176\u4ed6\u673a\u6784\u540d\u2018\u5927\u8fde\u7406\u5de5\u5927\u5b66\u2019\u3002<\/p>\n\n\n<p>5.\u4e2d\u6587\u91cc\u5e7f\u6cdb\u5b58\u5728\u7b80\u5316\u8868\u8fbe\uff0c\u5982\u5927\u5de5\u3001\u533b\u79d1\u5927\u7b49\u3002<\/p>\n\n\n<h2 class=\"wp-block-heading\">\u6570\u636e\u96c6\u6807\u6ce8\u65b9\u5f0f<\/h2>\n\n\n<p>BIO\u6807\u6ce8\u96c6\uff1a<\/p>\n\n\n<ul><li>B-PER\u3001I-PER\u4ee3\u8868\u4eba\u540d\u9996\u5b57\u3001\u4eba\u540d\u975e\u9996\u5b57\u3002<\/li><li>B-LOC\u3001I-LOC\u4ee3\u8868\u5730\u540d\u9996\u5b57\u3001\u5730\u540d\u975e\u9996\u5b57\u3002<\/li><li>B-ORG\u3001I-ORG\u4ee3\u8868\u7ec4\u7ec7\u673a\u6784\u540d\u9996\u5b57\u3001\u7ec4\u7ec7\u673a\u6784\u540d\u975e\u9996\u5b57\u3002<\/li><li>O\u4ee3\u8868\u8be5\u5b57\u4e0d\u5c5e\u4e8e\u547d\u540d\u5b9e\u4f53\u7684\u4e00\u90e8\u5206\u3002 <\/li><\/ul>\n\n\n<p>BIOSE\u6807\u6ce8\u96c6 \uff1a<\/p>\n\n\n<ul><li>B-PER\u3001I-PER\u3001E-PER\u4ee3\u8868\u4eba\u540d\u9996\u5b57\u3001\u4eba\u540d\u4e2d\u95f4\u5b57\u3001\u4eba\u540d\u672b\u5c3e\u5b57\u3002<\/li><li>B-LOC\u3001I-LOC\u3001E-PER\u4ee3\u8868\u5730\u540d\u9996\u5b57\u3001\u5730\u540d\u4e2d\u95f4\u5b57\u3001\u5730\u540d\u672b\u5c3e\u5b57\u3002<\/li><li>B-ORG\u3001I-ORG\u3001E-PER\u4ee3\u8868\u7ec4\u7ec7\u673a\u6784\u540d\u9996\u5b57\u3001\u7ec4\u7ec7\u673a\u6784\u540d\u4e2d\u95f4\u5b57\u3001 \u7ec4\u7ec7\u673a\u6784\u540d\u672b\u5c3e\u5b57\u3002<\/li><li>S-X\u4ee3\u8868\u5355\u5b57\u6784\u6210\u7684\u5b9e\u4f53\u3002<\/li><li>O\u4ee3\u8868\u8be5\u5b57\u4e0d\u5c5e\u4e8e\u547d\u540d\u5b9e\u4f53\u7684\u4e00\u90e8\u5206\u3002<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\u9690\u9a6c\u5c14\u79d1\u592b\u6a21\u578b\uff08HMM\uff09<\/h2>\n\n\n<p>\u9690\u9a6c\u5c14\u53ef\u592b\u6a21\u578b\uff08Hidden Markov Model\uff0cHMM\uff09\u63cf\u8ff0\u7531\u9690\u85cf\u7684\u9a6c\u5c14\u53ef\u592b\u94fe \u968f\u673a\u751f\u6210\u89c2\u6d4b\u5e8f\u5217\u7684\u8fc7\u7a0b\uff0c\u5c5e\u4e8e\u751f\u6210\u6a21\u578b\u3002<\/p>\n\n\n<p>HMM\u662f\u5173\u4e8e\u65f6\u5e8f\u7684\u6982\u7387\u6a21\u578b\uff0c\u63cf\u8ff0\u7531\u4e00\u4e2a\u9690\u85cf\u7684\u9a6c\u5c14\u53ef\u592b\u94fe\u968f\u673a\u751f\u6210\u4e0d\u53ef\u89c2 \u6d4b\u7684\u72b6\u6001\u968f\u673a\u5e8f\u5217\uff08\u72b6\u6001\u5e8f\u5217\uff09\uff0c\u518d\u7531\u5404\u4e2a\u72b6\u6001\u751f\u6210\u4e00\u4e2a\u89c2\u6d4b\u800c\u4ea7\u751f\u7684\u89c2\u6d4b\u968f \u673a\u5e8f\u5217\uff08\u89c2\u6d4b\u5e8f\u5217\uff09\u7684\u8fc7\u7a0b\uff0c\u9690\u9a6c\u5c14\u53ef\u592b\u6a21\u578b\u7531\u521d\u59cb\u72b6\u6001\u5206\u5e03\uff0c\u72b6\u6001\u8f6c\u79fb\u6982\u7387 \u77e9\u9635\u4ee5\u53ca\u89c2\u6d4b\u6982\u7387\u77e9\u9635\u6240\u786e\u5b9a\u3002<\/p>\n\n\n<p>NER\u672c\u8d28\u4e0a\u53ef\u4ee5\u770b\u6210\u662f\u4e00\u79cd\u5e8f\u5217\u6807\u6ce8\u95ee\u9898\uff08\u9884\u6d4b\u6bcf\u4e2a\u5b57\u7684BIOES\u6807\u8bb0\uff09\uff0c\u5728\u4f7f \u7528HMM\u89e3\u51b3NER\u8fd9\u79cd\u5e8f\u5217\u6807\u6ce8\u95ee\u9898\u7684\u65f6\u5019\uff0c\u6211\u4eec\u6240\u80fd\u89c2\u6d4b\u5230\u7684\u662f\u5b57\u7ec4\u6210\u7684\u5e8f \u5217\uff08\u89c2\u6d4b\u5e8f\u5217\uff09\uff0c\u89c2\u6d4b\u4e0d\u5230\u7684\u662f\u6bcf\u4e2a\u5b57\u5bf9\u5e94\u7684\u6807\u6ce8\uff08\u72b6\u6001\u5e8f\u5217\uff09\u3002 <\/p>\n\n\n<h2 class=\"wp-block-heading\">CRF(\u6761\u4ef6\u968f\u673a\u573a) <\/h2>\n\n\n<p>\u4e0a\u9762\u8bb2\u7684HMM\u6a21\u578b\u4e2d\u5b58\u5728\u4e24\u4e2a\u5047\u8bbe<\/p>\n\n\n<ul><li>\u8f93\u51fa\u89c2\u5bdf\u503c\u4e4b\u95f4\u4e25\u683c\u72ec\u7acb;<\/li><li> \u72b6\u6001\u8f6c\u79fb\u8fc7\u7a0b\u4e2d\u5f53\u524d\u72b6\u6001\u53ea\u4e0e\u524d\u4e00\u72b6\u6001\u6709\u5173; <\/li><\/ul>\n\n\n<p>\u4e5f\u5c31\u662f\u8bf4\uff0c\u5728\u547d\u540d\u5b9e\u4f53\u8bc6\u522b\u7684\u573a\u666f\u4e0b\uff0cHMM\u8ba4\u4e3a\u89c2\u6d4b\u5230\u7684\u53e5\u5b50\u4e2d\u7684\u6bcf \u4e2a\u5b57\u90fd\u662f\u76f8\u4e92\u72ec\u7acb\u7684\uff0c\u800c\u4e14\u5f53\u524d\u65f6\u523b\u7684\u6807\u6ce8\u53ea\u4e0e\u524d\u4e00\u65f6\u523b\u7684\u6807\u6ce8\u76f8\u5173\u3002 \u4f46\u5b9e\u9645\u4e0a\uff0c\u547d\u540d\u5b9e\u4f53\u8bc6\u522b\u5f80\u5f80\u9700\u8981\u66f4\u591a\u7684\u7279\u5f81\uff0c\u6bd4\u5982\u8bcd\u6027\uff0c\u8bcd\u7684\u4e0a\u4e0b\u6587 \u7b49\u7b49\uff0c\u540c\u65f6\u5f53\u524d\u65f6\u523b\u7684\u6807\u6ce8\u5e94\u8be5\u4e0e\u524d\u4e00\u65f6\u523b\u4ee5\u53ca\u540e\u4e00\u65f6\u523b\u7684\u6807\u6ce8\u90fd\u76f8\u5173 \u8054\u3002\u7531\u4e8e\u8fd9\u4e24\u4e2a\u5047\u8bbe\u7684\u5b58\u5728\uff0c\u663e\u7136HMM\u6a21\u578b\u5728\u89e3\u51b3\u547d\u540d\u5b9e\u4f53\u8bc6\u522b\u7684\u95ee \u9898\u4e0a\u662f\u5b58\u5728\u7f3a\u9677\u7684\u3002 <\/p>\n\n\n<p>\u800c\u6761\u4ef6\u968f\u673a\u573a\u5c31\u6ca1\u6709\u8fd9\u79cd\u95ee\u9898\uff0c\u5b83\u901a\u8fc7\u5f15\u5165\u81ea\u5b9a\u4e49\u7684\u7279\u5f81\u51fd\u6570\uff0c\u4e0d\u4ec5\u53ef \u4ee5\u8868\u8fbe\u89c2\u6d4b\u4e4b\u95f4\u7684\u4f9d\u8d56\uff0c\u8fd8\u53ef\u8868\u793a\u5f53\u524d\u89c2\u6d4b\u4e0e\u524d\u540e\u591a\u4e2a\u72b6\u6001\u4e4b\u95f4\u7684\u590d\u6742 \u4f9d\u8d56\uff0c\u53ef\u4ee5\u6709\u6548\u514b\u670dHMM\u6a21\u578b\u9762\u4e34\u7684\u95ee\u9898\u3002<\/p>\n\n\n<h2 class=\"wp-block-heading\">\u81a8\u80c0\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\uff08Dilation\uff09<\/h2>\n\n\n<p>\u81a8\u80c0\u5377\u79ef\u7684\u597d\u5904\u662f\uff1a<strong>\u4e0d\u505apolling\u635f\u5931<\/strong>\u4fe1\u606f\u7684\u60c5\u51b5\u4e0b\uff0c<strong>\u52a0\u5927\u4e86\u611f\u53d7\u91ce<\/strong>\uff0c\u8ba9\u6bcf\u4e2a\u5377\u79ef\u8f93\u51fa\u90fd\u5305\u542b\u8f83\u5927\u8303\u56f4\u7684\u4fe1\u606f\u3002\u5728<strong>\u56fe\u50cf\u9700\u8981\u5168\u5c40\u4fe1\u606f<\/strong>\u6216\u8005<strong>\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4e2d\u9700\u8981\u8f83\u957f\u7684\u5e8f\u5217\u4fe1\u606f\u4f9d\u8d56<\/strong>\u95ee\u9898\u4e2d\uff0c\u6709\u5f88\u597d\u7684\u5e94\u7528\u3002<\/p>\n\n\n<div class=\"wp-block-image\"><figure class=\"aligncenter\"><img loading=\"lazy\" decoding=\"async\" width=\"878\" height=\"389\" src=\"\/wp-content\/uploads\/2020\/09\/\u56fe\u7247.png\" alt=\"\" class=\"wp-image-3591\" srcset=\"http:\/\/www.sniper97.cn\/wp-content\/uploads\/2020\/09\/\u56fe\u7247.png 878w, http:\/\/www.sniper97.cn\/wp-content\/uploads\/2020\/09\/\u56fe\u7247-300x133.png 300w, http:\/\/www.sniper97.cn\/wp-content\/uploads\/2020\/09\/\u56fe\u7247-768x340.png 768w\" sizes=\"(max-width: 878px) 100vw, 878px\" \/><\/figure><\/div>\n\n\n<p>\u4ee3\u7801\u8be6\u89c1\uff1a<\/p>\n\n\n<p><a href=\"https:\/\/github.com\/Sniper970119\/nlp_base\">https:\/\/github.com\/Sniper970119\/nlp_base<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u547d\u540d\u5b9e\u4f53\u8bc6\u522b \u547d\u540d\u5b9e\u4f53\u8bc6\u522b\uff08Named Entity Recognition\uff0cNER\uff09\uff0c\u5b83\u662f\u81ea\u7136\u8bed\u8a00 [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"om_disable_all_campaigns":false,"_mi_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"categories":[14],"tags":[],"views":6842,"_links":{"self":[{"href":"http:\/\/www.sniper97.cn\/index.php\/wp-json\/wp\/v2\/posts\/3587"}],"collection":[{"href":"http:\/\/www.sniper97.cn\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.sniper97.cn\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.sniper97.cn\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.sniper97.cn\/index.php\/wp-json\/wp\/v2\/comments?post=3587"}],"version-history":[{"count":0,"href":"http:\/\/www.sniper97.cn\/index.php\/wp-json\/wp\/v2\/posts\/3587\/revisions"}],"wp:attachment":[{"href":"http:\/\/www.sniper97.cn\/index.php\/wp-json\/wp\/v2\/media?parent=3587"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.sniper97.cn\/index.php\/wp-json\/wp\/v2\/categories?post=3587"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.sniper97.cn\/index.php\/wp-json\/wp\/v2\/tags?post=3587"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}