[1]王仲昊,万相奎,李风从,等.多模型以动态权重相融合的词相似性分析[J].华侨大学学报(自然科学版),2021,42(1):121-127.[doi:10.11830/ISSN.1000-5013.202003035]
　WANG Zhonghao,WAN Xiangkui,LI Fengcong,et al.Word Similarity Analysis by Multi-Model Fusion WithDynamic Weight[J].Journal of Huaqiao University(Natural Science),2021,42(1):121-127.[doi:10.11830/ISSN.1000-5013.202003035]
点击复制

多模型以动态权重相融合的词相似性分析()
分享到：

《华侨大学学报（自然科学版）》[ISSN:1000-5013/CN:35-1079/N]

卷:
第42卷
期数:
2021年第1期
页码:
121-127
栏目:
出版日期:
2021-01-20

文章信息/Info

Title:
Word Similarity Analysis by Multi-Model Fusion WithDynamic Weight
文章编号:
1000-5013(2021)01-0121-07
作者:
王仲昊1;  万相奎1; 2; 3;  李风从1;  危竞1;  刘俊杰1
1. 湖北工业大学 太阳能高效利用及储能运行控制湖北省重点实验室, 湖北 武汉 430068;2. 湖北工业大学 太阳能高效利用湖北省协同创新中心, 湖北 武汉 430068;3. 湖北工业大学 湖北省电网智能控制与装备工程技术研究中心, 湖北 武汉 430068
Author(s):
WANG Zhonghao1;  WAN Xiangkui1; 2; 3;  LI Fengcong1; WEI Jing1;  LIU Junjie1
1. Key Laboratory of Solar Energy Efficient Utilization and Energy Storage Operation Control in Hubei Province, Hubei University of Technology, Wuhan 430068, China; 2. Solar Energy Efficient Use of Hubei Province Collaborative Innovation Center, Hubei Uni
关键词:
词相似性;  统计模型;  字典模型;  改进的多模型融合;  动态权重
Keywords:
word similarity;  statistical model;  dictionary model;  improved multi-model fusion;  dynamic weights
分类号:
TP391.1
DOI:
10.11830/ISSN.1000-5013.202003035
文献标志码:
A
摘要:
以NLPCC-ICCPOL 2016中文词语相似度比赛中的PKU-500数据集作为评价的参考标准,采用动态权重多模型融合的词相似性进行分析.将得到的斯皮尔曼等级相关系数0.568与NLPCC 2016第1名的结果相比,效果提高了9.6%.结果表明:基于动态权重改进的多模型融合方法,提高计算词相似性的准确率.
Abstract:
Taked the PKU-500 dataset in the NLPCC-ICCPOL 2016 Chinese word similarity competition as the reference standard for evaluation, used dynamic weights multi-model fusion to analyze word similarity, obtained a Spearman rank correlation coefficient of 0.568, which is 9.6% higher than the first place in the NLPCC-ICCPOL 2016. The results show that the improved multi-model fusion method based on dynamic weight improves the accuracy of calculating word similarity.

参考文献/References:

[1] LE Q,MIKOLOV T.Distributed representations of sentences and documents[C]//International Conference on Machine Learning.Pennsylvania:ICML,2014:1188-1196.
[2] HU Fei,LI Li,ZHANG Zili,et al.Emphasizing essential words for sentiment classification based on recurrent neural networks[J].计算机科学技术学报:英文版,2017,32(4):785-795.DOI:10.1007/s11390-017-1759-2.
[3] BOJANOWSKI P,GRAVE E,JOULIN A,et al.Enriching word vectors with subword information[J].Transactions of the Association for Computational Linguistics,2017,5:135-146.DOI:10.1162/tacl_a_00051.
[4] 李舟军,范宇,吴贤杰.面向自然语言处理的预训练技术研究综述[J].计算机科学,2020,47(3):162-173.DOI:10.11896/jsjkx.191000167.
[5] 梅家驹,竺一鸣,高蕴琦,等.同义词词林[M].上海:上海辞书出版社,1983.
[6] 马永起,韩德培,蒙立荣,等.基于How-net的词语语义相似度算法[J].计算机工程,2018,44(6):151-155.DOI:10.19678/j.issn.1000-3428.0047061.
[7] GUO Shaoru,GUAN Yong,LI Ru.Chinese word similarity computing base in combination strategy[C]//Proceedings of NLPCC 2016, Lecture Notes in Artificial Intelligence.Beijing:[s.n.],2016:744-752.DOI:10.1007/978-3-319-50496-4_67.
[8] WU Yunfang,LI Wei.Overview of the NLPCC-ICCPOL 2016 shared task: Chinese word similarity measurement[J].Lecture Notes in Artificial Intelligence,2016,10102:828-839.DOI:10.1007/978-3-319-50496-4_75.
[9] 吴军.数学之美[M].北京:人民邮电出版社,2014.
[10] MIKOLOV T,SUTSKEVER I,CHEN K,et al.Distributed representations of words and phrases and their compositionality[C]//Proceedings of the 27th Annual Conference on Neural Information processing Systems.New York:NIPS,2013:3111-3119.
[11] 廖胜兰,吉建民,俞畅,等.基于BERT模型与知识蒸馏的意图分类方法[J/OL].[2020-05-07] .https://doi.org/10.19678/j.issn.1000-3428.0057416.
[12] 陈慧,田大钢,冯成刚.多种算法对不同中文文本分类效果比较研究[J].软件导刊,2019,18(5):73-78.DOI:10.3969/j.issn.1003-0077.2019.03.005
[13] 朱靖雯,杨玉基,许斌,等.基于HowNet的语义表示学习[J].中文信息学报,2019,33(3):33-41.DOI:10.11907/rjdk.182489.
[14] 赵倩倩.词语相似度计算及其在语义选择限制知识获取中的应用研究[D].郑州:郑州大学,2018.
[15] 刘群,李素建.基于《知网》的词汇语义相似度计算[C]//第三届汉语词汇语义学研讨会论文集.台北:[s.n.],2002:59-76.
[16] 吕立辉,梁维薇,冉蜀阳.基于词林的词语相似度的度量[J].现代计算机(专业版),2013(1):3-6,9.DOI:10.3969/j.issn.1007-1423(z).2013.01.001.
[17] ZOU Yinfeng,OUYANG Chunping,LIU Yongbin,et al.A similarity algorithm based on the generality and individuality of words[C]//Proceedings of NLPCC 2016, Lecture Notes in Artificial Intelligence.Beijing:[s.n.],2016:549-588.DOI:10.1007/978-3-319-50496-4_48.

相似文献/References:

[1]黄丽玲,黄富贵.直线度误差统计模型识别的实验方法[J].华侨大学学报(自然科学版),2013,34(1):1.[doi:10.11830/ISSN.1000-5013.2013.01.0001]
　HUANG Li-ling,HUANG Fu-gui.Experimental Method of Straightness Error Statistical Model Identification[J].Journal of Huaqiao University(Natural Science),2013,34(1):1.[doi:10.11830/ISSN.1000-5013.2013.01.0001]

备注/Memo

备注/Memo:
收稿日期: 2020-03-29
通信作者: 万相奎(1976-),教授,博士,主要从事嵌入式系统设计与信号处理的研究.E-mail:wanxiangkui@163.com.
基金项目: 国家自然科学基金资助项目(61571182)
更新日期/Last Update: 2021-01-20