| { |
| "paper_id": "O17-3002", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T07:59:26.689134Z" |
| }, |
| "title": "On the Use of Neural Network Modeling Techniques for Spoken Document Retrieval", |
| "authors": [ |
| { |
| "first": "\u7f85\u5929\u5b8f", |
| "middle": [ |
| "\uf02a" |
| ], |
| "last": "\u3001\u9673\u6620\u6587", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "\uf02a", |
| "middle": [], |
| "last": "\u3001\u9673\u51a0\u5b87", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "\uf02b", |
| "middle": [], |
| "last": "\u3001\u738b\u65b0\u6c11", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "\uf023", |
| "middle": [], |
| "last": "\u3001\u9673\u67cf\u7433 \uf02a", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Tien-Hong", |
| "middle": [], |
| "last": "Lo", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "teinhonglo@ntnu.edu.tw" |
| }, |
| { |
| "first": "Ying-Wen", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Kuan-Yu", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "kychen@mail.ntust.edu.tw" |
| }, |
| { |
| "first": "Hsin-Min", |
| "middle": [], |
| "last": "Wang", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Berlin", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "berlin@ntnu.edu.tw" |
| } |
| ], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "Due to ever-increasing amounts of publicly available multimedia associated with speech information, spoken document retrieval (SDR) has been an active area of research that captures significant interest from both academic and industrial communities. Beyond the continuing effort in the development of robust indexing and effective retrieval methods to quantify the relevance degree between a pair of query and spoken document, how to accurately and efficiently model the query content plays a vital role for improving SDR performance. In view of this, we present in this paper a novel neural relevance-aware model (NRM) to infer an enhanced query representation, extricating the conventional time-consuming pseudo-relevance feedback (PRF) process. In addition, we incorporate the notion of query intent classification into our proposed NRM modeling framework to obtain more sophisticated query representations. Preliminary experiments conducted on the TDT-2 collection confirm the utility of our methods in relation to a few state-of-the-art ones.", |
| "pdf_parse": { |
| "paper_id": "O17-3002", |
| "_pdf_hash": "", |
| "abstract": [ |
| { |
| "text": "Due to ever-increasing amounts of publicly available multimedia associated with speech information, spoken document retrieval (SDR) has been an active area of research that captures significant interest from both academic and industrial communities. Beyond the continuing effort in the development of robust indexing and effective retrieval methods to quantify the relevance degree between a pair of query and spoken document, how to accurately and efficiently model the query content plays a vital role for improving SDR performance. In view of this, we present in this paper a novel neural relevance-aware model (NRM) to infer an enhanced query representation, extricating the conventional time-consuming pseudo-relevance feedback (PRF) process. In addition, we incorporate the notion of query intent classification into our proposed NRM modeling framework to obtain more sophisticated query representations. Preliminary experiments conducted on the TDT-2 collection confirm the utility of our methods in relation to a few state-of-the-art ones.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Abstract", |
| "sec_num": null |
| } |
| ], |
| "body_text": [ |
| { |
| "text": "\u4f34\u96a8\u8457\u7db2\u969b\u7db2\u8def\u7684\u767c\u5c55\u8207\u591a\u5a92\u9ad4\u8cc7\u8a0a\u7684\u5927\u91cf\u589e\u9577\uff0c\u5f71\u97f3\u7684\u700f\u89bd\u8207\u50b3\u905e\u4e5f\u9010\u6f38\u6210\u70ba\u6211\u5011\u7684 \u65e5\u5e38\u751f\u6d3b\u7684\u4e00\u90e8\u5206\u3002\u5728\u9019\u74b0\u5883\u4e0b\uff0c\u5982\u4f55\u5229\u7528\u8a9e\u97f3\u7684\u8cc7\u8a0a\uff0c\u5feb\u901f\u6aa2\u7d22\u7b26\u5408\u8cc7\u8a0a\u9700\u6c42\u7684\u5167\u5bb9\uff0c \u8b8a\u6210\u4e86\u4e00\u9805\u65b0\u8208\u7684\u9700\u6c42\u3002\u56e0\u6b64\uff0c\u5728\u904e\u53bb\u7684\u4e8c\u5341\u5e74 (Chelba, Hazen & Saraclar, 2008) (Lee & Chen, 2005) (Huang, Ma, Li & Wu, 2011) (Chen, Chen, Chen & Chen, 2012) \uff0c\u8a9e\u97f3\u6587\u4ef6\u6aa2\u7d22 \u6210\u70ba\u4e00\u500b\u5341\u5206\u6709\u9b45\u529b\u7684\u7814\u7a76\u4e3b\u984c\u3002\u5728\u8a9e\u97f3\u6587\u4ef6\u6aa2\u7d22\u7684\u4efb\u52d9\u4e0a\uff0c\u904e\u5f80\u6709\u8a31\u591a\u986f\u8457\u6210\u529f\u7684\u65b9 \u6cd5\uff0c\u5982\u5411\u91cf\u7a7a\u9593\u6a21\u578b(Vector Space Model) (Salton, Wong & Yang, 1975) \u3001Okapi BM25 model (Jones, Walker & Robertson, 2000) \uff0c\u4ee5\u53ca\u4e3b\u984c\u6a21\u578b(Topic Model) (Blei, Ng & Jordan, 2003) \u7b49\u3002\u53e6\u4e00\u65b9\u9762\uff0c\u5c07\u7d71\u8a08\u5f0f\u8a9e\u8a00\u6a21\u578b(Statistical Language Model)\u61c9\u7528\u5728\u6587\u5b57\u6aa2\u7d22 (Information Retrieval)\u548c\u8a9e\u97f3\u6587\u4ef6\u6aa2\u7d22\uff0c\u5728\u6aa2\u7d22\u4efb\u52d9\u4e0a\u53d6\u5f97\u4e86\u5d84\u65b0\u7684\u7a81\u7834 (Ponte & Croft, 1998) (Song & Croft, 1999) (Croft & Lafferty, 2003) \uff0c\u56e0\u6b64\u5438\u5f15\u4e86\u4e0d\u5c11\u7814\u7a76\u8005\u7684\u76ee\u5149\u3002\u5728\u9019 \u6a23\u7684\u6982\u5ff5\u4e0b\uff0c\u67e5\u8a62\u5c0d\u6bcf\u500b\u6587\u4ef6\u8a08\u7b97\u4f3c\u7136\u6a5f\u7387\u5f8c\u4f5c\u6392\u540d\uff0c\u6211\u5011\u7a31\u9019\u6a23\u7684\u6392\u5e8f\u65b9\u6cd5\u70ba\u67e5\u8a62\u4f3c \u7136\u6e2c\u91cf(Query Likelihood Model Measure, QLM) (Manning, Raghavan & Schutze, 2008) \u3002\u53e6 \u4e00\u500b\u77e5\u540d\u7684\u8a55\u4f30\u65b9\u5f0f\u70ba KL \u6563\u5ea6\u6e2c\u91cf(Kullback-Leibler Divergence Measure, KLM) (Zhai & Lafferty, 2001 )\uff0c\u5c07\u67e5\u8a62\u8207\u6587\u4ef6\u7686\u8868\u793a\u70ba\u55ae\u5143\u8a9e\u6cd5\u7684\u8a9e\u8a00\u6a21\u578b(Unigram Language Model)\uff0c \u67e5\u8a62\u8207\u6587\u4ef6\u7684\u76f8\u4f3c\u7a0b\u5ea6\u5373\u70ba\u5169\u500b\u6a5f\u7387\u5206\u4f48\u7684\u6563\u5ea6\u8ddd\u96e2(Divergence Distance)\u3002 \u6700\u8fd1\uff0c\u96a8\u8457\u6df1\u5c64\u985e\u795e\u7d93\u7db2\u8def\u67b6\u69cb\u7684\u6d41\u884c\uff0c\u9019\u985e\u7684\u65b9\u6cd5\u4e5f\u88ab\u5927\u91cf\u61c9\u7528\u5728\u6aa2\u7d22\u7684\u4efb\u52d9\u4e0a\u3002 \u4e3b\u8981\u7684\u7814\u7a76\u65b9\u5411\u70ba\u5229\u7528\u4e0d\u540c\u7db2\u8def\u67b6\u69cb\u8207\u8a13\u7df4\u65b9\u6cd5\uff0c\u4ee5\u6b64\u4f86\u5b78\u7fd2\u67e5\u8a62\u8207\u6587\u4ef6\u9593\u7684\u76f8\u4f3c\u95dc\u4fc2 (Guo, Fan, Ai & Croft, 2016) (Mitra, Diaz & Craswell, 2017) (Mikolov, Sutskever, Chen, Corrado & Dean, 2013 )\u548c\u9023\u7e8c\u8a5e\u888b (Continuous bag-of-words) )\u3002\u5f8c\u8005\u7684\u65b9\u6cd5\u6709\u5206\u4f48\u5f0f \u5167 \u5b58 \u6a21 \u578b (the distributed memory model) (Le & Mikolov, 2014 ) \u3001 \u5206 \u4f48 \u5f0f \u8a5e \u888b \u6a21 \u578b (distributed bag-of-words model) (Le & Mikolov, 2014) (Chen, Lee, Wang, Chen & Chen, 2014) \u548c\u672c\u8cea\u5411\u91cf\u6a21\u578b(essence vector model) )\u7b49\u3002 \u70ba\u4e86\u63d0\u5347\u6aa2\u7d22\u6548\u80fd\uff0c\u904e\u5f80\u6709\u8a31\u591a\u65b9\u6cd5\u5617\u8a66\u731c\u6e2c\u4f7f\u7528\u8005\u610f\u5411\u9032\u884c\u67e5\u8a62\u5206\u985e (Shen et al., 2006 )\uff0c\u7d66\u4e88\u5c0d\u4e0d\u540c\u985e\u5225\u6709\u8208\u8da3\u7684\u4f7f\u7528\u8005\u63d0\u4f9b\u66f4\u70ba\u7cbe\u78ba\u7684\u7d50\u679c\u3002\u7136\u800c\uff0c\u5982\u540c\u50b3\u7d71\u6587\u5b57\u6587\u4ef6 \u6aa2\u7d22\uff0c\u8a9e\u97f3\u6587\u4ef6\u6aa2\u7d22\u4e5f\u9762\u81e8\u4e86\u67e5\u8a62\u904e\u65bc\u7c21\u77ed\u8a9e\u610f\u4e0d\u6e05\uff0c\u4e14\u6703\u96a8\u8457\u6642\u9593\u63a8\u79fb\u6539\u8b8a\u8a9e\u53e5\u7684\u610f \u601d\uff0c\u96e3\u4ee5\u8868\u9054\u51fa\u4f7f\u7528\u8005\u7684\u8cc7\u8a0a\u9700\u6c42\uff0c\u56e0\u6b64\u67e5\u8a62\u5206\u985e\u5f80\u5f80\u6bd4\u6587\u4ef6\u5206\u985e\u56f0\u96e3\u3002\u91dd\u5c0d\u67e5\u8a62\u8a9e\u610f \u4e0d\u8db3\u7684\u554f\u984c\uff0c\u67e5\u8a62\u5206\u985e\u88ab\u5b9a\u7fa9\u70ba\u5c0d\u4f7f\u7528\u8005\u884c\u70ba\u7684\u5efa\u6a21\uff0c\u5176\u4e00\u4fbf\u662f\u8c50\u5bcc\u8a9e\u610f\u8868\u793a\u7684\u4efb\u52d9\uff0c \u7a31\u70ba\u64f4\u589e\u67e5\u8a62(Query Expansion)\uff0c\u64f4\u589e\u67e5\u8a62\u4e3b\u8981\u53ef\u4ee5\u5206\u70ba\u5169\u500b\u985e\u5225\uff0c\u7b2c\u4e00\u7a2e\u70ba\u5f15\u7528\u5916\u90e8 \u7684\u8cc7\u6e90(\u5982 Wikipedia \u6216 WordNet)\u5206\u6790\u8a9e\u610f\uff0c\u9032\u4e00\u6b65\u64f4\u5c55\u539f\u59cb\u7684\u67e5\u8a62\uff0c\u5176\u4e2d\u7684\u8a9e\u610f\u95dc\u4fc2\u5305 \u62ec\u540c\u7fa9\u8a5e\u3001\u53cd\u7fa9\u8a5e\u3001\u591a\u7fa9\u8a5e\u7b49\uff1b\u7b2c\u4e8c\u7a2e\u70ba\u5206\u6790\u67e5\u8a62\u7684\u56de\u994b\uff0c\u7d66\u4e88\u4e00\u500b\u67e5\u8a62\uff0c\u8ffd\u8e64\u4f7f\u7528\u8005 \u9ede\u64ca\u7684\u6587\u4ef6\uff0c\u4e26\u4ee5\u6b64\u505a\u5206\u985e\uff0c\u4ee5\u6c42\u7b2c\u4e8c\u6b21\u66f4\u70ba\u7cbe\u78ba\u7684\u7d50\u679c\uff0c\u8207\u975e\u76e3\u7763\u5f0f\u7684\u6e96\u76f8\u95dc\u56de\u994b (Pseudo-Relevance Feedback) (Zhai & Lafferty, 2001) (Lavrenko & Croft, 2001) \u7684\u7cbe\u795e\u76f8\u4f3c\u3002\u7b2c\u4e00\u500b\u65b9\u6cd5\u9700\u8981\u8f03\u8907\u96dc\u7684\u81ea\u7136\u8a9e\u8a00\u8655\u7406\u6280\u8853\uff0c\u5982 \u8a9e\u610f\u8868\u5fb5\u548c\u63a8\u8ad6\u3002\u7b2c\u4e8c\u7a2e\u65b9\u6cd5\u5247\u8f03\u70ba\u7c21\u55ae\uff0c\u53ea\u9700\u8981\u53d6\u5f97\u524d\u5e7e\u7bc7\u6587\u7ae0\u505a\u5206\u6790\uff0c\u518d\u9069\u7576\u5730\u8207 \u539f\u59cb\u67e5\u8a62\u505a\u7d50\u5408\u5373\u53ef\u3002\u4e14\u56e0\u70ba\u5206\u6790\u7684\u8cc7\u6599\u50c5\u70ba\u56de\u994b\u7684\u6587\u4ef6\uff0c\u6e96\u76f8\u95dc\u56de\u994b\u4e0d\u9700\u8981\u984d\u5916\u7684\u8a9e \u6599\u5eab\u4f86\u5b78\u7fd2\u3002\u4ee5\u4e0b\u70ba\u5e7e\u500b\u77e5\u540d\u7684\u6e96\u76f8\u95dc\u56de\u994b (Manning et al., 2008) ", |
| "cite_spans": [ |
| { |
| "start": 101, |
| "end": 133, |
| "text": "(Chelba, Hazen & Saraclar, 2008)", |
| "ref_id": "BIBREF2" |
| }, |
| { |
| "start": 134, |
| "end": 152, |
| "text": "(Lee & Chen, 2005)", |
| "ref_id": "BIBREF21" |
| }, |
| { |
| "start": 153, |
| "end": 179, |
| "text": "(Huang, Ma, Li & Wu, 2011)", |
| "ref_id": "BIBREF15" |
| }, |
| { |
| "start": 180, |
| "end": 211, |
| "text": "(Chen, Chen, Chen & Chen, 2012)", |
| "ref_id": "BIBREF3" |
| }, |
| { |
| "start": 289, |
| "end": 316, |
| "text": "(Salton, Wong & Yang, 1975)", |
| "ref_id": "BIBREF29" |
| }, |
| { |
| "start": 335, |
| "end": 368, |
| "text": "(Jones, Walker & Robertson, 2000)", |
| "ref_id": "BIBREF16" |
| }, |
| { |
| "start": 390, |
| "end": 415, |
| "text": "(Blei, Ng & Jordan, 2003)", |
| "ref_id": "BIBREF1" |
| }, |
| { |
| "start": 513, |
| "end": 534, |
| "text": "(Ponte & Croft, 1998)", |
| "ref_id": "BIBREF28" |
| }, |
| { |
| "start": 535, |
| "end": 555, |
| "text": "(Song & Croft, 1999)", |
| "ref_id": "BIBREF31" |
| }, |
| { |
| "start": 556, |
| "end": 580, |
| "text": "(Croft & Lafferty, 2003)", |
| "ref_id": "BIBREF10" |
| }, |
| { |
| "start": 679, |
| "end": 714, |
| "text": "(Manning, Raghavan & Schutze, 2008)", |
| "ref_id": "BIBREF23" |
| }, |
| { |
| "start": 779, |
| "end": 801, |
| "text": "(Zhai & Lafferty, 2001", |
| "ref_id": "BIBREF32" |
| }, |
| { |
| "start": 972, |
| "end": 1000, |
| "text": "(Guo, Fan, Ai & Croft, 2016)", |
| "ref_id": "BIBREF13" |
| }, |
| { |
| "start": 1001, |
| "end": 1031, |
| "text": "(Mitra, Diaz & Craswell, 2017)", |
| "ref_id": "BIBREF27" |
| }, |
| { |
| "start": 1032, |
| "end": 1079, |
| "text": "(Mikolov, Sutskever, Chen, Corrado & Dean, 2013", |
| "ref_id": "BIBREF26" |
| }, |
| { |
| "start": 1164, |
| "end": 1183, |
| "text": "(Le & Mikolov, 2014", |
| "ref_id": "BIBREF20" |
| }, |
| { |
| "start": 1235, |
| "end": 1255, |
| "text": "(Le & Mikolov, 2014)", |
| "ref_id": "BIBREF20" |
| }, |
| { |
| "start": 1256, |
| "end": 1292, |
| "text": "(Chen, Lee, Wang, Chen & Chen, 2014)", |
| "ref_id": "BIBREF4" |
| }, |
| { |
| "start": 1359, |
| "end": 1377, |
| "text": "(Shen et al., 2006", |
| "ref_id": "BIBREF30" |
| }, |
| { |
| "start": 1737, |
| "end": 1760, |
| "text": "(Zhai & Lafferty, 2001)", |
| "ref_id": "BIBREF32" |
| }, |
| { |
| "start": 1761, |
| "end": 1785, |
| "text": "(Lavrenko & Croft, 2001)", |
| "ref_id": "BIBREF18" |
| }, |
| { |
| "start": 1912, |
| "end": 1934, |
| "text": "(Manning et al., 2008)", |
| "ref_id": "BIBREF23" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "\u7dd2\u8ad6 (INTRODUCTION)", |
| "sec_num": "1." |
| }, |
| { |
| "text": "EQUATION", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [ |
| { |
| "start": 0, |
| "end": 8, |
| "text": "EQUATION", |
| "ref_id": "EQREF", |
| "raw_str": "RM \u5047\u8a2d\u6bcf\u4e00\u500b\u67e5\u8a62 \u7686\u6709\u4e00\u500b\u76f8\u95dc\u985e\u5225 \uff0c\u4e14\u6bcf\u500b\u8207\u67e5\u8a62 \u76f8\u95dc\u7684\u6587\u4ef6\u7686\u7531\u76f8\u95dc\u985e\u5225 \u4e2d\u7522\u751f\u3002\u4e0d\u5e78\u7684\u662f\uff0c\u6211\u5011\u4e0d\u6703\u77e5\u9053\u67e5\u8a62 \u7684\u76f8\u95dc\u985e\u5225 \uff0c\u56e0\u6b64\u4f5c\u70ba\u66ff\u4ee3\uff0c\u6211\u5011\u6703\u5c07\u524d N \u7bc7\u56de\u994b\u7684\u6587\u7ae0 \u7576\u4f5c\u76f8\u95dc\u6587\u7ae0\uff0c\u4e26\u4e14\u5229\u7528\u6e96\u76f8\u95dc\u56de\u994b\u8fd1\u4f3c\u771f\u5be6\u7684\u76f8\u95dc\u985e\u5225 \u3002\u5c0d\u61c9\u67e5\u8a62 \u7684\u76f8\u95dc\u6027\u6a21\u578b\u53ef\u5229\u7528\u4ee5\u4e0b\u516c\u5f0f\u8a08\u7b97\uff1a | \u2211 | \u220f \u2208 \u2208 \u2211 \u220f \u2208 \u2208 ,", |
| "eq_num": "(1)" |
| } |
| ], |
| "section": "\u7dd2\u8ad6 (INTRODUCTION)", |
| "sec_num": "1." |
| }, |
| { |
| "text": "EQUATION", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [ |
| { |
| "start": 0, |
| "end": 8, |
| "text": "EQUATION", |
| "ref_id": "EQREF", |
| "raw_str": "\u5176\u4e2d \u70ba\u6587\u4ef6\u7684\u7522\u751f\u6a5f\u7387\u3002\u7531\u65bc\u6211\u5011\u6c92\u6709\u5c0d\u65bc\u6587\u4ef6\u7684\u5148\u9a57\u77e5\u8b58\uff0c\u56e0\u6b64\u6c7a\u5b9a\u6a5f\u7387\u7684\u65b9\u5f0f \u662f\u7528\u5747\u52fb\u5206\u4f48(Uniform Distribution)\u4f86\u5be6\u73fe\u3002\u53e6\u4e00\u500b\u8a9e\u8a00\u6a21\u578b | \u5247\u662f\u5229\u7528\u6700\u5927\u4f3c\u7136 \u4f30\u8a08\u7684\u65b9\u5f0f\uff0c\u5229\u7528\u6587\u4ef6\u7684\u8a5e\u983b\u548c\u8a72\u6587\u4ef6\u8207\u67e5\u8a62\u76f8\u4f3c\u7a0b\u5ea6\uff0c\u8a08\u7b97\u51fa\u6bcf\u500b\u8a5e\u51fa\u73fe\u5728\u6587\u4ef6 \u7684 \u6a5f\u7387\uff0c\u5176\u9918\u7b26\u865f\u4ee5\u6b64\u985e\u63a8\u3002 2.2 \u7c21\u6613\u6df7\u5408\u6a21\u578b(Simple Mixture Model) \u53e6\u4e00\u500b\u4f30\u6e2c\u67e5\u8a62\u8a9e\u8a00\u6a21\u578b\u7684\u89c0\u9ede\u662f SMM\u3002SMM \u5047\u8a2d\u5728\u56de\u994b\u6587\u4ef6 \u88e1\u8a5e\u51fa\u73fe\u6a5f\u7387\u4e0d\u662f\u7531 \u55ae\u4e00\u6a21\u578b\u4f30\u6e2c\uff0c\u800c\u662f\u4f86\u81ea\u5169\u90e8\u4efd\u6df7\u5408\u800c\u6210\u7684\u8a9e\u8a00\u6a21\u578b\uff0c\u7b2c\u4e00\u90e8\u4efd\u662f\u5c08\u5c6c\u65bc\u8a72\u67e5\u8a62 \u7684\u7279 \u6b8a\u8a5e\u7684\u4e3b\u984c\u6a21\u578b | \uff1b\u7b2c\u4e8c\u90e8\u4efd\u662f\u5ee3\u6cdb\u51fa\u73fe\u5728\u5404\u500b\u6587\u4ef6\u4e2d\u7684\u80cc\u666f\u8a9e\u8a00\u6a21\u578b \u3002 \u5982\u6b64\u4e00\u4f86\uff0c\u4fbf\u53ef\u900f\u904e\u5206\u6790\u56de\u994b\u6587\u4ef6 \uff0c\u6700\u5927\u5316 SMM \u7684\u4f3c\u7136\u6a5f\u7387\uff0c\u4e26\u6c42\u5f97 | \uff0c \u4ee5\u4e0b\u70ba\u4f30\u6e2c\u6642\u4f7f\u7528\u7684\u640d\u5931\u51fd\u6578(loss function)\uff1a \u220f \u220f \u2022 | 1 \u2022 , \u2208 \u2208 ,", |
| "eq_num": "(2)" |
| } |
| ], |
| "section": "\u7dd2\u8ad6 (INTRODUCTION)", |
| "sec_num": "1." |
| }, |
| { |
| "text": "EQUATION", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [ |
| { |
| "start": 0, |
| "end": 8, |
| "text": "EQUATION", |
| "ref_id": "EQREF", |
| "raw_str": "| \u2022 \u2022 | 1 \u2022 | ,", |
| "eq_num": "(3)" |
| } |
| ], |
| "section": "\u7dd2\u8ad6 (INTRODUCTION)", |
| "sec_num": "1." |
| } |
| ], |
| "back_matter": [], |
| "bib_entries": { |
| "BIBREF0": { |
| "ref_id": "b0", |
| "title": "Modern information retrieval: the concepts and technology behind search", |
| "authors": [ |
| { |
| "first": "R", |
| "middle": [], |
| "last": "Baeza-Yates", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Ribeiro-Neto", |
| "suffix": "" |
| } |
| ], |
| "year": 2011, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Baeza-Yates, R. & Ribeiro-Neto, B. (2011). Modern information retrieval: the concepts and technology behind search. Boston, MA: Addison-Wesley Professional.", |
| "links": null |
| }, |
| "BIBREF1": { |
| "ref_id": "b1", |
| "title": "Latent Dirichlet Allocation", |
| "authors": [ |
| { |
| "first": "D", |
| "middle": [ |
| "M" |
| ], |
| "last": "Blei", |
| "suffix": "" |
| }, |
| { |
| "first": "A", |
| "middle": [ |
| "Y" |
| ], |
| "last": "Ng", |
| "suffix": "" |
| }, |
| { |
| "first": "M", |
| "middle": [ |
| "I" |
| ], |
| "last": "Jordan", |
| "suffix": "" |
| } |
| ], |
| "year": 2003, |
| "venue": "Journal of Machine Learning Research", |
| "volume": "3", |
| "issue": "4-5", |
| "pages": "993--1022", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Blei, D. M., Ng, A. Y. & Jordan, M. I. (2003). Latent Dirichlet Allocation. Journal of Machine Learning Research, 3(4-5), 993-1022.", |
| "links": null |
| }, |
| "BIBREF2": { |
| "ref_id": "b2", |
| "title": "Retrieval and browsing of spoken content", |
| "authors": [ |
| { |
| "first": "C", |
| "middle": [], |
| "last": "Chelba", |
| "suffix": "" |
| }, |
| { |
| "first": "T", |
| "middle": [ |
| "J" |
| ], |
| "last": "Hazen", |
| "suffix": "" |
| }, |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Saraclar", |
| "suffix": "" |
| } |
| ], |
| "year": 2008, |
| "venue": "IEEE Signal Processing Magazine", |
| "volume": "25", |
| "issue": "3", |
| "pages": "39--49", |
| "other_ids": { |
| "DOI": [ |
| "10.1109/MSP.2008.917992" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Chelba, C., Hazen, T. J. & Saraclar, M. (2008). Retrieval and browsing of spoken content. IEEE Signal Processing Magazine, 25(3), 39-49. doi: 10.1109/MSP.2008.917992", |
| "links": null |
| }, |
| "BIBREF3": { |
| "ref_id": "b3", |
| "title": "Spoken document retrieval with unsupervised query modeling techniques", |
| "authors": [ |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "K.-Y", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "P.-N", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "Y.-W", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| } |
| ], |
| "year": 2012, |
| "venue": "IEEE Transactions on Audio, Speech, and Language Processing", |
| "volume": "20", |
| "issue": "9", |
| "pages": "2602--2612", |
| "other_ids": { |
| "DOI": [ |
| "10.1109/TASL.2012.2208628" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Chen, B., Chen, K.-Y., Chen, P.-N. & Chen, Y.-W. (2012). Spoken document retrieval with unsupervised query modeling techniques. IEEE Transactions on Audio, Speech, and Language Processing, 20(9), 2602-2612. doi: 10.1109/TASL.2012.2208628", |
| "links": null |
| }, |
| "BIBREF4": { |
| "ref_id": "b4", |
| "title": "I-vector Based Language Modeling for Spoken Document Retrieval", |
| "authors": [ |
| { |
| "first": "K", |
| "middle": [ |
| "Y" |
| ], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [ |
| "S" |
| ], |
| "last": "Lee", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [ |
| "M" |
| ], |
| "last": "Wang", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [ |
| "H" |
| ], |
| "last": "Chen", |
| "suffix": "" |
| } |
| ], |
| "year": 2014, |
| "venue": "Proceedings of 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)", |
| "volume": "", |
| "issue": "", |
| "pages": "7083--7088", |
| "other_ids": { |
| "DOI": [ |
| "10.1109/ICASSP.2014.6854974" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Chen, K. Y., Lee, H. S., Wang, H. M., Chen, B. & Chen, H. H. (2014). I-vector Based Language Modeling for Spoken Document Retrieval. In Proceedings of 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 7083-7088. doi: 10.1109/ICASSP.2014.6854974", |
| "links": null |
| }, |
| "BIBREF5": { |
| "ref_id": "b5", |
| "title": "Learning to Distill: The Essence Vector Modeling Framework", |
| "authors": [ |
| { |
| "first": "K", |
| "middle": [ |
| "Y" |
| ], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [ |
| "H" |
| ], |
| "last": "Liu", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [ |
| "M" |
| ], |
| "last": "Wang", |
| "suffix": "" |
| } |
| ], |
| "year": 2016, |
| "venue": "Proceedings of the 26th International Conference on Computational Linguistics", |
| "volume": "", |
| "issue": "", |
| "pages": "358--368", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Chen, K. Y., Liu, S. H., Chen, B. & Wang, H. M. (2016). Learning to Distill: The Essence Vector Modeling Framework. In Proceedings of the 26th International Conference on Computational Linguistics (COLING 2016), 358-368.", |
| "links": null |
| }, |
| "BIBREF6": { |
| "ref_id": "b6", |
| "title": "A locality-preserving essence vector modeling framework for spoken document retrieval", |
| "authors": [ |
| { |
| "first": "K", |
| "middle": [ |
| "Y" |
| ], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [ |
| "H" |
| ], |
| "last": "Liu", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [ |
| "M" |
| ], |
| "last": "Wang", |
| "suffix": "" |
| } |
| ], |
| "year": 2017, |
| "venue": "Proceedings of ICASSP 2017", |
| "volume": "", |
| "issue": "", |
| "pages": "5665--5669", |
| "other_ids": { |
| "DOI": [ |
| "10.1109/ICASSP.2017.7953241" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Chen, K. Y., Liu, S. H., Chen, B. & Wang, H. M. (2017). A locality-preserving essence vector modeling framework for spoken document retrieval. In Proceedings of ICASSP 2017, 5665-5669. doi: 10.1109/ICASSP.2017.7953241", |
| "links": null |
| }, |
| "BIBREF8": { |
| "ref_id": "b8", |
| "title": "Exploring the use of unsupervised query modeling techniques for speech recognition and summarization", |
| "authors": [ |
| { |
| "first": "K", |
| "middle": [ |
| "Y" |
| ], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [ |
| "H" |
| ], |
| "last": "Liu", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [ |
| "M" |
| ], |
| "last": "Wang", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [ |
| "H" |
| ], |
| "last": "Chen", |
| "suffix": "" |
| } |
| ], |
| "year": 2016, |
| "venue": "Speech Communication", |
| "volume": "80", |
| "issue": "", |
| "pages": "49--59", |
| "other_ids": { |
| "DOI": [ |
| "10.1016/j.specom.2016.03.006" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Chen, K. Y., Liu, S. H., Chen, B., Wang, H. M. & Chen, H. H. (2016). Exploring the use of unsupervised query modeling techniques for speech recognition and summarization. Speech Communication, 80, 49-59. doi: 10.1016/j.specom.2016.03.006", |
| "links": null |
| }, |
| "BIBREF9": { |
| "ref_id": "b9", |
| "title": "Exploring the use of significant words language modeling for spoken document retrieval", |
| "authors": [ |
| { |
| "first": "Y", |
| "middle": [ |
| "W" |
| ], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "K", |
| "middle": [ |
| "Y" |
| ], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [ |
| "M" |
| ], |
| "last": "Wang", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| } |
| ], |
| "year": 2017, |
| "venue": "Proceedings of INTERSPEECH 2017", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": { |
| "DOI": [ |
| "10.21437/Interspeech.2017-612" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Chen, Y. W., Chen, K. Y., Wang, H. M. & Chen, B. (2017). Exploring the use of significant words language modeling for spoken document retrieval. In Proceedings of INTERSPEECH 2017. doi: 10.21437/Interspeech.2017-612", |
| "links": null |
| }, |
| "BIBREF10": { |
| "ref_id": "b10", |
| "title": "Language modeling for information retrieval", |
| "authors": [ |
| { |
| "first": "W", |
| "middle": [ |
| "B" |
| ], |
| "last": "Croft", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Lafferty", |
| "suffix": "" |
| } |
| ], |
| "year": 2003, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": { |
| "DOI": [ |
| "10.1007/978-94-017-0171-6" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Croft, W. B. & Lafferty, J. (2003). Language modeling for information retrieval. Dordrecht, the Netherlands : Kluwer Academic Publishers. doi: 10.1007/978-94-017-0171-6", |
| "links": null |
| }, |
| "BIBREF11": { |
| "ref_id": "b11", |
| "title": "Luhn revisited: significant words language models", |
| "authors": [ |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Dehghani", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [], |
| "last": "Azarbonyad", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Kamps", |
| "suffix": "" |
| }, |
| { |
| "first": "D", |
| "middle": [], |
| "last": "Hiemstra", |
| "suffix": "" |
| }, |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Marx", |
| "suffix": "" |
| } |
| ], |
| "year": 2016, |
| "venue": "Proceedings of the 25th ACM International on Conference on Information and Knowledge Management (CIKM '16", |
| "volume": "", |
| "issue": "", |
| "pages": "1301--1310", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/2983323.2983814" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Dehghani, M., Azarbonyad, H., Kamps, J., Hiemstra, D. & Marx, M. (2016). Luhn revisited: significant words language models. In Proceedings of the 25th ACM International on Conference on Information and Knowledge Management (CIKM '16), 1301-1310. doi: 10.1145/2983323.2983814", |
| "links": null |
| }, |
| "BIBREF12": { |
| "ref_id": "b12", |
| "title": "Maximum likelihood from incomplete data via the em algorithm", |
| "authors": [ |
| { |
| "first": "A", |
| "middle": [ |
| "P" |
| ], |
| "last": "Dempster", |
| "suffix": "" |
| }, |
| { |
| "first": "N", |
| "middle": [ |
| "M" |
| ], |
| "last": "Laird", |
| "suffix": "" |
| }, |
| { |
| "first": "D", |
| "middle": [ |
| "B" |
| ], |
| "last": "Rubin", |
| "suffix": "" |
| } |
| ], |
| "year": 1977, |
| "venue": "Journal of the Royal Statistical Society", |
| "volume": "39", |
| "issue": "1", |
| "pages": "1--38", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Dempster, A. P., Laird, N. M. & Rubin, D. B. (1977). Maximum likelihood from incomplete data via the em algorithm. Journal of the Royal Statistical Society, 39(1), 1-38.", |
| "links": null |
| }, |
| "BIBREF13": { |
| "ref_id": "b13", |
| "title": "A deep relevance matching model for ad-hoc retrieval", |
| "authors": [ |
| { |
| "first": "J.-F", |
| "middle": [], |
| "last": "Guo", |
| "suffix": "" |
| }, |
| { |
| "first": "Y", |
| "middle": [], |
| "last": "Fan", |
| "suffix": "" |
| }, |
| { |
| "first": "Q", |
| "middle": [], |
| "last": "Ai", |
| "suffix": "" |
| }, |
| { |
| "first": "W", |
| "middle": [ |
| "B" |
| ], |
| "last": "Croft", |
| "suffix": "" |
| } |
| ], |
| "year": 2016, |
| "venue": "Proceedings of CIKM '16", |
| "volume": "", |
| "issue": "", |
| "pages": "55--64", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/2983323.2983769" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Guo, J.-F., Fan, Y., Ai, Q. & Croft, W. B. (2016). A deep relevance matching model for ad-hoc retrieval. In Proceedings of CIKM '16, 55-64. doi: 10.1145/2983323.2983769", |
| "links": null |
| }, |
| "BIBREF14": { |
| "ref_id": "b14", |
| "title": "Learning deep structured semantic models for web search using clickthrough data", |
| "authors": [ |
| { |
| "first": "P", |
| "middle": [ |
| "S" |
| ], |
| "last": "Huang", |
| "suffix": "" |
| }, |
| { |
| "first": "X", |
| "middle": [], |
| "last": "He", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Gao", |
| "suffix": "" |
| }, |
| { |
| "first": "L", |
| "middle": [], |
| "last": "Deng", |
| "suffix": "" |
| }, |
| { |
| "first": "A", |
| "middle": [], |
| "last": "Acero", |
| "suffix": "" |
| }, |
| { |
| "first": "L", |
| "middle": [], |
| "last": "Heck", |
| "suffix": "" |
| } |
| ], |
| "year": 2013, |
| "venue": "Proceedings of CIKM '13", |
| "volume": "", |
| "issue": "", |
| "pages": "2333--2338", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/2505515.2505665" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Huang, P. S., He, X., Gao, J., Deng, L., Acero, A. & Heck, L. (2013). Learning deep structured semantic models for web search using clickthrough data. In Proceedings of CIKM '13, 2333-2338. doi: 10.1145/2505515.2505665", |
| "links": null |
| }, |
| "BIBREF15": { |
| "ref_id": "b15", |
| "title": "Speech indexing using semantic context inference", |
| "authors": [ |
| { |
| "first": "C", |
| "middle": [ |
| "L" |
| ], |
| "last": "Huang", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Ma", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [], |
| "last": "Li", |
| "suffix": "" |
| }, |
| { |
| "first": "C.-H", |
| "middle": [], |
| "last": "Wu", |
| "suffix": "" |
| } |
| ], |
| "year": 2011, |
| "venue": "Proceedings of INTERSPEECH 2011", |
| "volume": "", |
| "issue": "", |
| "pages": "717--720", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Huang, C. L., Ma, B., Li, H. & Wu, C.-H. (2011). Speech indexing using semantic context inference. In Proceedings of INTERSPEECH 2011, 717-720.", |
| "links": null |
| }, |
| "BIBREF16": { |
| "ref_id": "b16", |
| "title": "A probabilistic model of information retrieval: development and comparative experiments (Parts 1 and 2). Information Processing and Management", |
| "authors": [ |
| { |
| "first": "K", |
| "middle": [ |
| "S" |
| ], |
| "last": "Jones", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [], |
| "last": "Walker", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [ |
| "E" |
| ], |
| "last": "Robertson", |
| "suffix": "" |
| } |
| ], |
| "year": 2000, |
| "venue": "", |
| "volume": "36", |
| "issue": "", |
| "pages": "779--840", |
| "other_ids": { |
| "DOI": [ |
| "10.1016/S0306-4573(00)00015-7" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Jones, K. S., Walker, S. & Robertson, S. E. (2000). A probabilistic model of information retrieval: development and comparative experiments (Parts 1 and 2). Information Processing and Management, 36(6), 779-840. doi: 10.1016/S0306-4573(00)00015-7", |
| "links": null |
| }, |
| "BIBREF17": { |
| "ref_id": "b17", |
| "title": "ADAM: A method for stochastic optimization", |
| "authors": [ |
| { |
| "first": "D", |
| "middle": [], |
| "last": "Kingma", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Ba", |
| "suffix": "" |
| } |
| ], |
| "year": 2015, |
| "venue": "Proceedings of ICLR", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Kingma, D. & Ba, J. (2015). ADAM: A method for stochastic optimization. In Proceedings of ICLR 2015.", |
| "links": null |
| }, |
| "BIBREF18": { |
| "ref_id": "b18", |
| "title": "Relevance based language models", |
| "authors": [ |
| { |
| "first": "V", |
| "middle": [], |
| "last": "Lavrenko", |
| "suffix": "" |
| }, |
| { |
| "first": "W", |
| "middle": [ |
| "B" |
| ], |
| "last": "Croft", |
| "suffix": "" |
| } |
| ], |
| "year": 2001, |
| "venue": "Proceedings of the 24th annual international ACM SIGIR conference on Research and development in information retrieval (SIGIR '01)", |
| "volume": "", |
| "issue": "", |
| "pages": "120--127", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/383952.383972" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Lavrenko, V. & Croft, W. B. (2001). Relevance based language models. In Proceedings of the 24th annual international ACM SIGIR conference on Research and development in information retrieval (SIGIR '01), 120-127. doi: 10.1145/383952.383972", |
| "links": null |
| }, |
| "BIBREF19": { |
| "ref_id": "b19", |
| "title": "Project of Topic Detection and Tracking", |
| "authors": [], |
| "year": 2000, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Linguistic Data Consortium. (2000). Project of Topic Detection and Tracking.", |
| "links": null |
| }, |
| "BIBREF20": { |
| "ref_id": "b20", |
| "title": "Distributed representations of sentences and documents", |
| "authors": [ |
| { |
| "first": "Q", |
| "middle": [], |
| "last": "Le", |
| "suffix": "" |
| }, |
| { |
| "first": "T", |
| "middle": [], |
| "last": "Mikolov", |
| "suffix": "" |
| } |
| ], |
| "year": 2014, |
| "venue": "Proceedings of ICML '14", |
| "volume": "", |
| "issue": "", |
| "pages": "1188--1196", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Le, Q. & Mikolov, T. (2014). Distributed representations of sentences and documents. In Proceedings of ICML '14, 1188-1196.", |
| "links": null |
| }, |
| "BIBREF21": { |
| "ref_id": "b21", |
| "title": "Spoken document understanding and organization", |
| "authors": [ |
| { |
| "first": "L", |
| "middle": [ |
| "S" |
| ], |
| "last": "Lee", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| } |
| ], |
| "year": 2005, |
| "venue": "IEEE Signal Processing Magazine", |
| "volume": "22", |
| "issue": "5", |
| "pages": "42--60", |
| "other_ids": { |
| "DOI": [ |
| "10.1109/MSP.2005.1511823" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Lee, L. S. & Chen, B. (2005). Spoken document understanding and organization. IEEE Signal Processing Magazine, 22(5), 42-60. doi: 10.1109/MSP.2005.1511823", |
| "links": null |
| }, |
| "BIBREF22": { |
| "ref_id": "b22", |
| "title": "The automatic creation of literature abstracts", |
| "authors": [ |
| { |
| "first": "H", |
| "middle": [ |
| "P" |
| ], |
| "last": "Luhn", |
| "suffix": "" |
| } |
| ], |
| "year": 1958, |
| "venue": "IBM Journal of research and development", |
| "volume": "2", |
| "issue": "2", |
| "pages": "159--165", |
| "other_ids": { |
| "DOI": [ |
| "10.1147/rd.22.0159" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Luhn, H. P. (1958). The automatic creation of literature abstracts. IBM Journal of research and development, 2(2), 159-165. doi: 10.1147/rd.22.0159", |
| "links": null |
| }, |
| "BIBREF23": { |
| "ref_id": "b23", |
| "title": "Introduction to Information Retrieval", |
| "authors": [ |
| { |
| "first": "C", |
| "middle": [ |
| "D" |
| ], |
| "last": "Manning", |
| "suffix": "" |
| }, |
| { |
| "first": "P", |
| "middle": [], |
| "last": "Raghavan", |
| "suffix": "" |
| }, |
| { |
| "first": "H", |
| "middle": [], |
| "last": "Schutze", |
| "suffix": "" |
| } |
| ], |
| "year": 2008, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Manning, C. D., Raghavan, P. & Schutze, H. (2008). Introduction to Information Retrieval. New York, NY: Cambridge University Press.", |
| "links": null |
| }, |
| "BIBREF24": { |
| "ref_id": "b24", |
| "title": "Mandarin-English information (MEI): investigating translingual speech retrieval", |
| "authors": [ |
| { |
| "first": "H", |
| "middle": [], |
| "last": "Meng", |
| "suffix": "" |
| }, |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [], |
| "last": "Khudanpur", |
| "suffix": "" |
| }, |
| { |
| "first": "G.-A", |
| "middle": [], |
| "last": "Levow", |
| "suffix": "" |
| }, |
| { |
| "first": "W.-K", |
| "middle": [], |
| "last": "Lo", |
| "suffix": "" |
| }, |
| { |
| "first": "D", |
| "middle": [], |
| "last": "Oard", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "\u2026wang", |
| "suffix": "" |
| } |
| ], |
| "year": 2004, |
| "venue": "Computer Speech and Language", |
| "volume": "18", |
| "issue": "2", |
| "pages": "163--179", |
| "other_ids": { |
| "DOI": [ |
| "10.1016/j.csl.2003.09.003" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Meng, H., Chen, B., Khudanpur, S., Levow, G.-A., Lo, W.-K., Oard, D., \u2026Wang, J. (2004). Mandarin-English information (MEI): investigating translingual speech retrieval. Computer Speech and Language, 18(2), 163-179. doi: 10.1016/j.csl.2003.09.003", |
| "links": null |
| }, |
| "BIBREF25": { |
| "ref_id": "b25", |
| "title": "Efficient estimation of word representations in vector space", |
| "authors": [ |
| { |
| "first": "T", |
| "middle": [], |
| "last": "Mikolov", |
| "suffix": "" |
| }, |
| { |
| "first": "K", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "G", |
| "middle": [], |
| "last": "Corrado", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Dean", |
| "suffix": "" |
| } |
| ], |
| "year": 2013, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": { |
| "arXiv": [ |
| "arXiv:1301.3781" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Mikolov, T., Chen, K., Corrado, G. & Dean, J. (2013). Efficient estimation of word representations in vector space. Retrieved from arXiv:1301.3781.", |
| "links": null |
| }, |
| "BIBREF26": { |
| "ref_id": "b26", |
| "title": "Distributed representations of words and phrases and their compositionality", |
| "authors": [ |
| { |
| "first": "T", |
| "middle": [], |
| "last": "Mikolov", |
| "suffix": "" |
| }, |
| { |
| "first": "I", |
| "middle": [], |
| "last": "Sutskever", |
| "suffix": "" |
| }, |
| { |
| "first": "K", |
| "middle": [], |
| "last": "Chen", |
| "suffix": "" |
| }, |
| { |
| "first": "G", |
| "middle": [], |
| "last": "Corrado", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Dean", |
| "suffix": "" |
| } |
| ], |
| "year": 2013, |
| "venue": "Proceedings of NIPS '13", |
| "volume": "", |
| "issue": "", |
| "pages": "3111--3119", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Mikolov, T., Sutskever, I., Chen, K., Corrado, G. & Dean, J. (2013). Distributed representations of words and phrases and their compositionality. In Proceedings of NIPS '13, 3111-3119.", |
| "links": null |
| }, |
| "BIBREF27": { |
| "ref_id": "b27", |
| "title": "Learning to match using local and distributed representations of text for web search", |
| "authors": [ |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Mitra", |
| "suffix": "" |
| }, |
| { |
| "first": "F", |
| "middle": [], |
| "last": "Diaz", |
| "suffix": "" |
| }, |
| { |
| "first": "N", |
| "middle": [], |
| "last": "Craswell", |
| "suffix": "" |
| } |
| ], |
| "year": 2017, |
| "venue": "Proceedings of WWW '17", |
| "volume": "", |
| "issue": "", |
| "pages": "1291--1299", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/3038912.3052579" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Mitra, B., Diaz, F. & Craswell, N. (2017). Learning to match using local and distributed representations of text for web search. In Proceedings of WWW '17, 1291-1299. doi: 10.1145/3038912.3052579", |
| "links": null |
| }, |
| "BIBREF28": { |
| "ref_id": "b28", |
| "title": "A language modeling approach to information retrieval", |
| "authors": [ |
| { |
| "first": "J", |
| "middle": [ |
| "M" |
| ], |
| "last": "Ponte", |
| "suffix": "" |
| }, |
| { |
| "first": "W", |
| "middle": [ |
| "B" |
| ], |
| "last": "Croft", |
| "suffix": "" |
| } |
| ], |
| "year": 1998, |
| "venue": "Proceedings of SIGIR '98", |
| "volume": "", |
| "issue": "", |
| "pages": "275--281", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/290941.291008" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Ponte, J. M. & Croft, W. B. (1998). A language modeling approach to information retrieval. In Proceedings of SIGIR '98, 275-281. doi: 10.1145/290941.291008", |
| "links": null |
| }, |
| "BIBREF29": { |
| "ref_id": "b29", |
| "title": "A vector space model for automatic indexing", |
| "authors": [ |
| { |
| "first": "G", |
| "middle": [], |
| "last": "Salton", |
| "suffix": "" |
| }, |
| { |
| "first": "A", |
| "middle": [], |
| "last": "Wong", |
| "suffix": "" |
| }, |
| { |
| "first": "C", |
| "middle": [ |
| "S" |
| ], |
| "last": "Yang", |
| "suffix": "" |
| } |
| ], |
| "year": 1975, |
| "venue": "Communications of the ACM", |
| "volume": "18", |
| "issue": "11", |
| "pages": "613--620", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/361219.361220" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Salton, G., Wong, A. & Yang, C. S. (1975). A vector space model for automatic indexing. Communications of the ACM, 18(11), 613-620. doi: 10.1145/361219.361220", |
| "links": null |
| }, |
| "BIBREF30": { |
| "ref_id": "b30", |
| "title": "Query enrichment for web-query classification", |
| "authors": [ |
| { |
| "first": "D", |
| "middle": [], |
| "last": "Shen", |
| "suffix": "" |
| }, |
| { |
| "first": "R", |
| "middle": [], |
| "last": "Pan", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [ |
| "T" |
| ], |
| "last": "Sun", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [ |
| "J" |
| ], |
| "last": "Pan", |
| "suffix": "" |
| }, |
| { |
| "first": "K", |
| "middle": [], |
| "last": "Wu", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Yin", |
| "suffix": "" |
| }, |
| { |
| "first": "Q", |
| "middle": [], |
| "last": "Yang", |
| "suffix": "" |
| } |
| ], |
| "year": 2006, |
| "venue": "ACM Transactions on Information Systems (TOIS)", |
| "volume": "24", |
| "issue": "3", |
| "pages": "320--352", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/1165774.1165776" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Shen, D., Pan, R., Sun, J. T., Pan, J. J., Wu, K., Yin, J., & Yang, Q. (2006). Query enrichment for web-query classification. ACM Transactions on Information Systems (TOIS), 24(3), 320-352. doi: 10.1145/1165774.1165776", |
| "links": null |
| }, |
| "BIBREF31": { |
| "ref_id": "b31", |
| "title": "A general language model for information retrieval", |
| "authors": [ |
| { |
| "first": "F", |
| "middle": [], |
| "last": "Song", |
| "suffix": "" |
| }, |
| { |
| "first": "W", |
| "middle": [ |
| "B" |
| ], |
| "last": "Croft", |
| "suffix": "" |
| } |
| ], |
| "year": 1999, |
| "venue": "Proceedings of CIKM '99", |
| "volume": "", |
| "issue": "", |
| "pages": "316--321", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/319950.320022" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Song, F. & Croft, W. B. (1999). A general language model for information retrieval. In Proceedings of CIKM '99, 316-321. doi: 10.1145/319950.320022", |
| "links": null |
| }, |
| "BIBREF32": { |
| "ref_id": "b32", |
| "title": "Model-based feedback in the language modeling approach to information retrieval", |
| "authors": [ |
| { |
| "first": "C", |
| "middle": [], |
| "last": "Zhai", |
| "suffix": "" |
| }, |
| { |
| "first": "J", |
| "middle": [], |
| "last": "Lafferty", |
| "suffix": "" |
| } |
| ], |
| "year": 2001, |
| "venue": "Proceedings of CIKM '01", |
| "volume": "", |
| "issue": "", |
| "pages": "403--410", |
| "other_ids": { |
| "DOI": [ |
| "10.1145/502585.502654" |
| ] |
| }, |
| "num": null, |
| "urls": [], |
| "raw_text": "Zhai, C. & Lafferty, J. (2001). Model-based feedback in the language modeling approach to information retrieval. In Proceedings of CIKM '01, 403-410. doi: 10.1145/502585.502654", |
| "links": null |
| } |
| }, |
| "ref_entries": { |
| "FIGREF0": { |
| "type_str": "figure", |
| "num": null, |
| "uris": null, |
| "text": "\u96e2\u7dda\u6642\u5229\u7528\u9ede\u64ca\u8cc7\u8a0a\u8a13\u7df4\u7684 NRM \u6a21\u578b\u4e4b\u6aa2\u7d22\u7d50\u679c [Figure 3. Retrieval results of the NRM offline trained on clickresults of the NRM trained with pseudo relevance feedback.]" |
| }, |
| "TABREF4": { |
| "type_str": "table", |
| "num": null, |
| "text": "", |
| "content": "<table><tr><td/><td colspan=\"4\">\u8a9e\u97f3\u6587\u4ef6\u6aa2\u7d22\u4f7f\u7528\u985e\u795e\u7d93\u7db2\u8def\u6280\u8853 \u8a9e\u97f3\u6587\u4ef6\u6aa2\u7d22\u4f7f\u7528\u985e\u795e\u7d93\u7db2\u8def\u6280\u8853</td><td/><td colspan=\"2\">11 \u7f85\u5929\u5b8f \u7b49 13</td></tr><tr><td colspan=\"8\">\u63a5\u8457\u662f\u6211\u5011\u5047\u8a2d\u7684\u7b2c\u4e8c\u500b\u60c5\u5883\uff0c\u67e5\u8a62\u8207\u76f8\u95dc\u6587\u4ef6\u7684\u95dc\u4fc2\u662f\u672a\u77e5\u3002\u5728\u9019\u6b21\u7684\u5be6\u9a57\u4e2d\uff0c \u5728 NRM \u7684\u67b6\u69cb\u4e0b\u8a13\u7df4\u6a21\u578b\uff0c\u640d\u5931\u503c\u53ef\u4ee5\u964d\u5f97\u6bd4\u539f\u5148\u96c6\u7fa4\u8f03\u5c11\u7684\u72c0\u6cc1\u4e0b\u66f4\u4f4e(\u5982\u5206\u6210\u56db\u500b</td></tr><tr><td colspan=\"8\">\u6211\u5011\u67e5\u8a62\u7d50\u679c\u7684\u524d 10 \u7bc7\u7576\u4f5c\u76f8\u95dc\u6587\u4ef6\uff0c\u56e0\u6b64\u76f8\u95dc\u6587\u4ef6\u672a\u5fc5\u771f\u7684\u6b63\u78ba\u3002\u6211\u5011\u53ef\u4ee5\u5f9e\u7d50\u679c\u4e2d \u96c6\u7fa4\u7684\u7b2c\u4e09\u500b\u96c6\u7fa4)\u3002\u56e0\u6b64\u7528\u65bc\u6355\u6349\u4f7f\u7528\u8005\u610f\u5411(Intent)\u7684\u8a13\u7df4\u65b9\u5f0f\uff0c\u8b93\u7db2\u8def\u5f97\u5230\u66f4\u591a\u8cc7\u8a0a</td></tr><tr><td colspan=\"8\">\u89c0\u5bdf\u5230\u5e7e\u500b\u73fe\u8c61\u3002\u9996\u5148\uff0c\u8207\u8868 2 \u4e0d\u540c\uff0cLPEV \u5728\u9019\u6b21\u7684\u5be6\u9a57\u4e2d\uff0c\u8868\u73fe\u52dd\u904e DSSM\u3002\u56e0\u70ba \u7684\u60c5\u6cc1\u4e0b\uff0c\u80fd\u66f4\u6709\u6548\u5730\u8a13\u7df4 NRM\u3002\u7e3d\u7d50\u5148\u524d\u6240\u505a\u7684\u5be6\u9a57\uff0c\u4ee5\u4e0a\u7a2e\u7a2e\u5be6\u9a57\u63ed\u793a\u4e86\u4e00\u4e9b\u8a0a\u606f\uff0c</td></tr><tr><td colspan=\"8\">LPEV \u7684\u76ee\u6a19\u8207 DSSM \u4e0d\u540c\uff0c\u4e5f\u56e0\u6b64\u8fa8\u8b58\u932f\u8aa4\u7684\u76f8\u95dc\u6587\u4ef6\u7684\u5f71\u97ff\u4e5f\u8f03\u5c0f\uff0c\u8868\u73fe\u8f03\u70ba\u7a69\u5065\u3002 \u4e0d\u8ad6\u662f\u5728\u8a9e\u97f3\u6587\u4ef6\u6aa2\u7d22\u6216\u8cc7\u8a0a\u6aa2\u7d22\u7684\u9818\u57df\u4e0a\uff0c\u6211\u5011\u63d0\u51fa\u7684 NRM \u90fd\u53ef\u5f97\u5230\u66f4\u597d\u7684\u7d50\u679c\uff0c</td></tr><tr><td colspan=\"8\">\u5728\u8a9e\u97f3\u6587\u4ef6\u7684\u90e8\u5206\uff0cSWM \u7684\u8868\u73fe\u4f9d\u820a\u53ef\u4ee5\u52dd\u904e SMM \u548c RM\uff0c\u50c5\u5728\u624b\u5beb\u6587\u672c\u7684\u90e8\u5206\u7565\u5fae \u8207\u6700\u65b0\u7a4e\u7684\u8a9e\u8a00\u6a21\u578b\u6bd4\u8f03\u8d77\u4f86\u6beb\u4e0d\u905c\u8272\u3002</td></tr><tr><td colspan=\"8\">\u4e0b\u964d\u3002\u6700\u5f8c\uff0c\u6574\u9ad4\u4f86\u770b\uff0c\u6bd4\u8f03\u5716 3 \u8207\u5716 4\uff0c\u9019\u6b21\u7684\u5be6\u9a57\u666e\u904d\u8868\u73fe\u8f03\u5dee\uff0c\u53ef\u4ee5\u89c0\u5bdf\u5230\u6b63\u78ba \u7b54\u6848\u5c0d\u65bc\u4ee5\u4e0a\u5e7e\u500b\u65b9\u6cd5\u7684\u5f71\u97ff\uff0c\u4ee5 NRM \u7684\u7d50\u679c\u4f86\u770b\uff0c\u53ef\u4ee5\u767c\u73fe NRM \u975e\u5e38\u4f9d\u8cf4\u76f8\u95dc\u6587\u4ef6 5. \u7d50\u8ad6 (CONCLUSIONS)</td></tr><tr><td colspan=\"8\">\u7684\u4fe1\u606f\u662f\u5426\u6b63\u78ba\u3002\u6700\u5f8c\uff0c\u5373\u4f7f\u662f\u5728\u975e\u76e3\u7763\u7684\u74b0\u5883\u4e0b\uff0cNRM \u4ecd\u7136\u53ef\u5728\u5404\u7a2e\u60c5\u5883\u4e0b\uff0c\u8868\u73fe\u6bd4 \u5728\u9019\u7bc7\u8ad6\u6587\u4e2d\uff0c\u6211\u5011\u63d0\u51fa\u4e00\u500b\u5efa\u7acb\u5728 NRM \u4e4b\u4e0a\u7684\u67e5\u8a62\u610f\u5411\u63a2\u7d22\u65b9\u6cd5\u3002\u5728\u8a9e\u97f3\u6587\u4ef6\u6aa2\u7d22</td></tr><tr><td colspan=\"8\">LPEV\u3001DSSM \u512a\u7570\uff0c\u518d\u4e00\u6b21\u8b49\u660e NRM \u7684\u512a\u79c0\u4e4b\u8655\u3002 \u7684\u4efb\u52d9\u4e2d\uff0cNRM \u7684\u65b9\u6cd5\u80fd\u5920\u5728\u4e0d\u9700\u8981\u7e41\u96dc\u7684\u6e96\u76f8\u95dc\u56de\u994b\u7684\u8655\u7406\u4e0b\uff0c\u5f97\u5230\u4e00\u500b\u66f4\u6709\u9451\u5225\u529b</td></tr><tr><td colspan=\"8\">\u7684\u67e5\u8a62\u8a9e\u8a00\u6a21\u578b\uff0c\u5927\u5e45\u63d0\u5347\u6aa2\u7d22\u7684\u6548\u80fd\u3002\u5be6\u9a57\u7684\u7d50\u679c\u4e5f\u8b49\u5be6\uff0c\u9019\u6a23\u7684\u91cd\u69cb\u67e5\u8a62\u8a9e\u8a00\u6a21\u578b \u7684\u6280\u8853\uff0c\u6bd4\u8d77\u904e\u5f80\u7684\u76f8\u95dc\u6280\u8853\uff0c\u6aa2\u7d22\u6548\u80fd\u7686\u80fd\u7a69\u5b9a\u7684\u52dd\u51fa\u3002\u5118\u7ba1\u521d\u6b65\u52a0\u5165\u67e5\u8a62\u610f\u5411\u7684\u7d50 \u8868 2. \u9ede\u64ca\u8cc7\u8a0a \u6e96\u76f8\u95dc\u56de\u994b \u679c\u4e0d\u76e1\u7406\u60f3\uff0c\u4f46\u5be6\u9a57\u7d50\u679c\u63ed\u793a\u4ecd\u6709\u8a13\u7df4 NRM \u67e5\u8a62\u610f\u5411\u7684\u53ef\u80fd\u3002\u672a\u4f86\u7684\u5de5\u4f5c\uff0c\u6211\u5011\u5e0c\u671b</td></tr><tr><td colspan=\"8\">\u6587\u5b57\u6587\u4ef6 \u80fd\u5920\u5617\u8a66\u4e00\u4e9b\u66f4\u8907\u96dc\u7684\u985e\u795e\u7d93\u7db2\u8def\u6a21\u578b(\u5982\u647a\u7a4d\u795e\u7d93\u7db2\u8def(CNN)\u3001\u905e\u6b78\u795e\u7d93\u7db2\u8def(RNN)\u7b49) \u8a9e\u97f3\u6587\u4ef6 \u6587\u5b57\u6587\u4ef6 \u8a9e\u97f3\u6587\u4ef6 \u4f86\u4f5c\u70ba NRM \u7684\u9aa8\u5e79\u3002\u6b64\u5916\uff0c\u6211\u5011\u4e5f\u5c07\u5617\u8a66\u52a0\u5165\u4e00\u4e9b\u65b0\u7684\u7279\u5fb5(\u5982\u53e5\u6cd5\u6216\u97fb\u5f8b)\uff0c\u89c0\u5bdf\u7db2\u8def</td></tr><tr><td colspan=\"8\">Long \u7372\u5f97\u66f4\u591a\u8cc7\u8a0a\u7684\u60c5\u6cc1\u4e0b\u80fd\u5426\u589e\u76ca\u5b78\u7fd2\u6548\u679c\u3002\u6700\u5f8c\u5247\u662f\u63d0\u4f9b\u66f4\u8907\u96dc\u7684\u67e5\u8a62\u610f\u5411\u65b9\u6cd5\uff0c\u4ee5\u6c42 Short Long Short Long Short Long Short</td></tr><tr><td>NRM SWM \u66f4\u70ba\u7d30\u7dfb\u7684\u67e5\u8a62\u7d50\u679c\u3002 0.730</td><td>0.563</td><td>0.686</td><td>0.547</td><td>0.648</td><td>0.467</td><td>0.589</td><td>0.449</td></tr><tr><td>NRM SWM)-2 0.690</td><td>0.571</td><td>0.670</td><td>0.544</td><td>0.636</td><td>0.475</td><td>0.593</td><td>0.470</td></tr><tr><td>NRM(SWM)-4 0.694</td><td>0.562-</td><td>0.669</td><td>0.545</td><td>0.628</td><td>0.462</td><td>0.583</td><td>0.434</td></tr><tr><td>NRM(SWM)-8 0.712</td><td>0.564-</td><td>0.672</td><td>0.547</td><td>0.632</td><td>0.463</td><td>0.593</td><td>0.437</td></tr><tr><td colspan=\"8\">\u5206\u7fa4\u7684\u5be6\u9a57\u90e8\u5206\uff0c\u6211\u5011\u5047\u8a2d\u67e5\u8a62\u4e4b\u9593\u662f\u6709\u4e0d\u540c\u985e\u5225\u7684\u95dc\u4fc2\uff0c\u56e0\u6b64\u5efa\u7acb\u5728\u539f\u5148 NRM \u7684\u57fa\u790e\u4e4b\u4e0a\uff0c\u5229\u7528\u7c21\u55ae\u7684\u5206\u7fa4\u6f14\u7b97\u6cd5\u5c07\u67e5\u8a62\u5206\u7fa4\uff0c\u4e0d\u540c\u7684\u985e\u5225\u5c31\u8a13\u7df4\u65b0\u7684 NRM \u6a21\u578b\uff0c \u671f\u671b\u80fd\u9054\u5230 NRM \u80fd\u5b78\u7fd2\u5230\u4e0d\u540c\u4e3b\u984c\u7684\u7279\u5fb5\uff0c\u9032\u4e00\u6b65\u63d0\u5347\u5b78\u7fd2\u7684\u6548\u679c\u3002\u9019\u88e1\u7684\u5be6\u9a57\uff0c\u6211 \u5011\u63a1\u7528\u7684\u662f\u5728\u7b2c\u4e00\u500b\u53ca\u7b2c\u4e8c\u500b\u5834\u666f\u4e0b\u7686\u8868\u73fe\u8f03\u70ba\u4eae\u773c\u7684 SWM \u65b9\u6cd5\u3002\u9996\u5148\uff0c\u6211\u5011\u5c07\u8a13\u7df4 \u5716 \u70ba\u4e86\u9032\u4e00\u6b65\u7814\u7a76\u9019\u6b21\u5206\u7fa4\u5c0d\u7d50\u679c\u7684\u5f71\u97ff\uff0c\u6211\u5011\u5c07\u7db2\u8def\u8a13\u7df4\u7684\u640d\u5931\u503c(loss)\u8996\u89ba\u5316\uff0c\u8a18</td></tr><tr><td colspan=\"8\">\u8cc7\u6599\u4e2d\u7684\u67e5\u8a62\u5206\u7fa4\uff0c\u4e26\u4f9d\u64da\u4e0d\u540c\u7684\u7fa4\u8a13\u7df4\u65b0\u7684 NRM\uff0c\u6e2c\u8a66\u7684\u9577(\u77ed)\u67e5\u8a62\u6703\u4f9d\u64da\u4e0d\u540c\u6b78\u5c6c \u9304\u5728\u5716 5\u3002\u7e31\u8ef8\u70ba\u640d\u5931\u503c\uff0c\u6a6b\u8ef8\u70ba\u8fed\u4ee3\u6b21\u6578\u3002\u4e0d\u540c\u7684\u7dda\u4ee3\u8868\u4e0d\u540c\u7db2\u8def\u8a13\u7df4\u6642\u7684\u5169\u500b\u503c\uff0c</td></tr><tr><td colspan=\"8\">\u7684\u7fa4\uff0c\u6c7a\u5b9a\u4f7f\u7528\u90a3\u500b NRM \u9810\u6e2c\u65b0\u7684\u67e5\u8a62\u8868\u793a\u5f0f\uff0c\u4e26\u4ee5\u6b64\u65b0\u7684\u6a21\u578b\u7d50\u5408\u820a\u6709\u9810\u6e2c\u51fa\u4f86\u7684 \u8a13\u7df4\u96c6(Training Set)\u548c\u9a57\u8b49\u96c6(Validation Set)\u7684\u640d\u5931\u503c\uff0c\u7531\u5de6\u5716\u5230\u53f3\u5716\uff0c\u5206\u5225\u662f\u4e0d\u5206\u7fa4\uff0c</td></tr><tr><td colspan=\"8\">\u6a21\u578b(\u5716 3 \u8207\u5716 4)\u7dda\u6027\u7d44\u5408\uff0c\u4ee5\u6b64\u505a\u70ba\u65b0\u7684\u67e5\u8a62\uff0c\u5be6\u9a57\u7d50\u679c\u5448\u73fe\u5728\u8868 2\u3002\u5f9e\u5be6\u9a57\u7d50\u679c\u53ef\u4ee5 \u4ee5\u53ca\u4e8c\u3001\u56db\u548c\u516b\u500b\u96c6\u7fa4\u7684\u8a13\u7df4\u7d50\u679c\uff0c\u6709\u4e9b\u96c6\u7fa4\u53ea\u5206\u5230\u4e00\u500b\u67e5\u8a62\uff0c\u96c6\u7fa4\u5927\u5c0f\u70ba\u4e00\u5c31\u4e0d\u5207\u6210</td></tr><tr><td colspan=\"8\">\u770b\u51fa\uff0c\u4e0d\u8ad6\u662f\u90a3\u4e00\u7a2e\u7684\u60c5\u5883\uff0c\u5728\u5206\u7fa4\u5f8c\u7684\u8a13\u7df4\u7d50\u679c\uff0c\u5927\u90e8\u5206\u7684\u6548\u80fd\u90fd\u662f\u4e0b\u964d\u6216\u6301\u5e73\uff0c\u5c11 \u8a13\u7df4\u96c6\u548c\u9a57\u8b49\u96c6\u3002\u5f9e\u4e0a\u5716\u4e2d\u53ef\u4ee5\u770b\u51fa\uff0c\u5206\u6210\u4e8c\u500b\u96c6\u7fa4\u7684\u8a13\u7df4\u904e\u7a0b\u8f03\u70ba\u7a69\u5b9a\uff0c\u8a13\u7df4\u96c6\u548c\u9a57</td></tr><tr><td colspan=\"8\">\u6578\u5e7e\u500b\u72c0\u6cc1\u4e0b\u8868\u73fe\u5f97\u6bd4\u539f\u5148\u7684\u7d50\u679c\u8f03\u597d\u3002\u5e73\u5747\u4f86\u770b\uff0c\u5206\u5169\u7fa4\u7684\u6548\u679c\u52dd\u904e\u56db\u7fa4\u8207\u516b\u7fa4\uff0c\u56db \u8b49\u96c6\u7684\u640d\u5931\u503c\u7686\u70ba\u7a69\u5b9a\u5730\u4e0b\u964d\u3002\u5206\u6210\u56db\u500b\u96c6\u7fa4\u5f8c\uff0c\u8a13\u7df4\u60c5\u6cc1\u6709\u4e9b\u4e0d\u540c\uff0c\u96d6\u7136\u8a13\u7df4\u96c6\u7684\u640d</td></tr><tr><td colspan=\"8\">\u7fa4\u7684\u6548\u80fd\u7d93\u5e38\u843d\u5728\u5206\u5169\u7fa4\u8207\u5206\u516b\u7fa4\u4e4b\u9593\uff0c\u5076\u800c\u5206\u516b\u7fa4\u7684\u6548\u679c\u6703\u662f\u6700\u4f73\u7684\uff0c\u4f46\u537b\u4e0d\u6703\u8d0f\u904e \u5931\u503c\u4ecd\u662f\u7a69\u5b9a\u4e0b\u964d\uff0c\u4f46\u9a57\u8b49\u96c6\u7684\u640d\u5931\u503c\u4e2d\u8868\u73fe\u8f03\u5dee\uff0c\u5c24\u5176\u662f\u7b2c\u96f6\u7fa4\u548c\u7b2c\u4e8c\u7fa4\u3002\u5206\u6210\u516b\u500b</td></tr><tr><td colspan=\"8\">\u592a\u591a\u3002\u9019\u6a23\u7684\u5be6\u9a57\u7d50\u679c\uff0c\u53ef\u4ee5\u8996\u70ba SWM \u662f\u8f03\u70ba\u8907\u96dc\u7684\u8a9e\u8a00\u6a21\u578b\uff0c\u6703\u6839\u64da\u4e0d\u540c\u7684\u67e5\u8a62\u6709 \u96c6\u7fa4\u5f8c\uff0c\u9019\u6a23\u7684\u73fe\u8c61\u5c31\u66f4\u70ba\u660e\u986f\uff0c\u8a13\u7df4\u96c6\u7684\u640d\u5931\u503c\u4f9d\u820a\u7a69\u5b9a\u4e0b\u964d\uff0c\u4f46\u9a57\u8b49\u96c6\u7684\u640d\u5931\u503c\u5247</td></tr><tr><td colspan=\"8\">\u4e0d\u540c\u7684\u5f71\u97ff\uff0c\u6240\u4ee5\u7576\u6211\u5011\u53ea\u5c07\u8a13\u7df4\u8cc7\u6599\u5206\u6210\u5169\u7fa4\uff0c\u5c0d\u500b\u5225\u8a13\u7df4\u7684\u7db2\u8def\u4f86\u8aaa\uff0c\u53ef\u4ee5\u5b78\u7fd2\u5230 \u6c92\u6709\u5448\u7a69\u5b9a\u7684\u4e0b\u964d\u66f2\u7dda\uff0c\u6709\u4e9b\u96c6\u7fa4\u7684\u640d\u5931\u503c\u751a\u81f3\u6709\u4e9b\u4e0a\u5347\uff0c\u53ef\u4ee5\u770b\u51fa\u660e\u986f\u5730\u904e\u5ea6\u8a13\u7df4</td></tr><tr><td colspan=\"8\">\u7684\u8a13\u7df4\u8cc7\u6599\u8f03\u70ba\u591a\u6a23\uff0c\u56e0\u6b64\u7db2\u8def\u8f03\u80fd\u5b78\u7fd2\u5230 SWM \u7684\u6e96\u76f8\u95dc\u56de\u994b\u7684\u7279\u6027\u3002\u53cd\u4e4b\uff0c\u5207\u5272\u8d8a (Overfitting)\u3002\u9664\u4e86\u904e\u591a\u7684\u96c6\u7fa4\u53ef\u80fd\u6703\u9020\u6210\u904e\u5ea6\u8a13\u7df4\u4ee5\u5916\uff0c\u6211\u5011\u4e5f\u89c0\u5bdf\u5230\u5206\u8d8a\u591a\u96c6\u7fa4\uff0c\u90e8</td></tr><tr><td colspan=\"8\">\u591a\u7684\u8cc7\u6599\u5f8c\uff0c\u8b93\u7db2\u8def\u7684\u6548\u80fd\u5247\u8b8a\u5f97\u8f03\u5dee\u3002\u5118\u7ba1\u5206\u5169\u7fa4\u7684\u6548\u679c\u666e\u904d\u52dd\u65bc\u5176\u4ed6\u7684\u8a2d\u5b9a\uff0c \u4f46\u6574 \u5206\u96c6\u7fa4\u7684\u8868\u73fe(\u8a13\u7df4\u96c6\u548c\u9a57\u8b49\u96c6)\u4e0b\u964d\u7684\u6bd4\u539f\u5148\u8f03\u5c11\u96c6\u7fa4\u7684\u591a(\u5982\u5206\u6210\u56db\u500b\u96c6\u7fa4\u7684\u7b2c\u4e09\u500b\u96c6</td></tr><tr><td colspan=\"8\">\u9ad4\u7684\u6548\u80fd\u6bd4\u8d77\u820a\u6709\u7684\u6a21\u578b\uff0c\u5927\u90e8\u5206\u4ecd\u662f\u6301\u5e73\u6216\u9000\u6b65\u3002 \u7fa4)\uff0c\u4f46\u9019\u6a23\u7684\u8868\u73fe\u4e26\u4e0d\u662f\u7a69\u5b9a\u7684\u7d50\u679c\uff0c\u53ef\u80fd\u662f\u539f\u5148\u5c07\u67e5\u8a62\u5206\u7fa4\u7684\u4f9d\u64da\u662f\u8a5e\u888b\u6a21\u578b\uff0c\u904e\u77ed</td></tr><tr><td colspan=\"8\">\u7684\u67e5\u8a62\u9020\u6210\u8a9e\u610f\u4e0d\u6e05\uff0c\u5c0e\u81f4\u5206\u7fa4\u6548\u679c\u4e0d\u5f70\uff0c\u90e8\u5206\u96c6\u7fa4\u6c92\u8a13\u7df4\u597d\uff0c\u9023\u5e36\u5f71\u97ff\u6574\u9ad4\u7684\u6210\u679c\u3002</td></tr><tr><td colspan=\"8\">\u5118\u7ba1\u9019\u6b21\u5be6\u9a57\u7684\u7d50\u679c\u4e0d\u5982\u9810\u671f\uff0c\u4f46\u6211\u5011\u4f9d\u820a\u767c\u73fe\uff0c\u900f\u904e\u7c21\u55ae\u7684\u67e5\u8a62\u5206\u985e\uff0c\u53ef\u4ee5\u66f4\u6709\u6548\u5730</td></tr></table>", |
| "html": null |
| } |
| } |
| } |
| } |
