A Transfer Method of Emotional Instances for Unbalanced Interactive Texts Based on Hyperplane Distance

doi:10.7652/xjtuxb201810001

您当前的位置：

首页 >

文章列表页 >

A Transfer Method of Emotional Instances for Unbalanced Interactive Texts Based on Hyperplane Distance

更新时间：2025-07-09

- A Transfer Method of Emotional Instances for Unbalanced Interactive Texts Based on Hyperplane Distance
- Vol. 52, Issue 10, Pages: 1-7(2018)
- 作者机构：
  
  1. 西安交通大学陕西省天地网技术重点实验室,西安,710049
  2. 西安交通大学电子与信息工程学院,西安,710049
- 作者简介：
- 基金信息：
- DOI：10.7652/xjtuxb201810001
  CLC： TP391.1
- Online First：10 October 2018，
  
  Published：2018
- 稿件说明：
移动端阅览
A Transfer Method of Emotional Instances for Unbalanced Interactive Texts Based on Hyperplane Distance[J]. 2018, 52(10): 1-7.
DOI：

A Transfer Method of Emotional Instances for Unbalanced Interactive Texts Based on Hyperplane Distance[J]. 2018, 52(10): 1-7. DOI： 10.7652/xjtuxb201810001.

摘要

针对非平衡交互文本少数类实例匮乏易导致训练的情感分类模型泛化性能差的问题

提出基于超平面距离的非平衡交互文本情感实例迁移方法。该方法将在少数类和多数类支持向量之间的源数据集实例作为待迁实例

并基于目标数据集上的分类超平面构造一个偏移超平面。依据最优信息效用原则基于待迁实例到偏移超平面的距离最短来筛选迁入的实例

同时通过调节迁入比例控制迁入实例规模生成合成数据集。实验结果表明:随着迁入实例增多

合成数据集对原始分布的偏离增大

所训练的序列最小优化算法(SMO)模型的泛化分类性能呈现先提升后降低的现象

类似于信息效用的Wundt曲线; 与SMOTE、Subsampling、Oversampling 3种数据层处理方法相比

所提方法训练的SMO、LibSVM、随机森林、代价敏感、CNN 5个分类模型在少数类识别F值上平均获得11%的增幅

且迁入比例最佳范围为20%～30%

在有效缓解非平衡特性的同时提高了少数类识别的泛化分类性能。

Abstract

A transfer method of emotional instances for unbalanced interactive texts is proposed based on hyperplane distance to focus the problem of poor generalization ability of sentiment classification models when they are trained on an unbalanced interactive text dataset that lacks of minority-class instances. The method uses instances of source dataset between support vectors of the minority class and the majority class as the transferrable instances

and constructs an offset hyperplane based on the classification hyperplane on the target dataset. The method uses the principle of optimal information utility to select the transfer instances based on the shortest distance between the instances and the offset hyperplane

and adopts the migration ratio to control the size of the transfer instances and to generate a synthetic dataset. Experiment results show that when transfer instances increase

the deviation of the synthetic dataset from the original distribution increases

and the generalized classification performance of the trained SMO model rises at the beginning and then decreases after it reaches its maximum

which is similar to the Wundt curve of the information utility. Comparisons with three data layer processing methods(SMOTE

Subsampling and Oversampling)show that five classification models(SMO

LibSVM

random forest

cost sensitive and CNN)trained by the proposed method obtain an average increase of 11% in the F-value of recognizing the minority class

and the optimal range of the migration ratio is [20%

30%]. It is concluded that the proposed method effectively alleviates the unbalanced characteristics and raises the generalized classification performance of the minority class.

关键词

Keywords

references

TIAN Feng, WU Fan, CHAO Kuo-Ming, et al. A topic sentence-based instance transfer method for imbalanced sentiment classification of Chinese product reviews [J]. Electronic Commerce Research and Applications, 2016, 16(3): 66-76.

Al-STOUHI S, REDDY C K. Transfer learning for class imbalance problems with inadequate data [J]. Knowledge Information Systems, 2016, 48(1): 201-228.

CORTES C, VAPNIK V. Support-vector network [J]. Machine Learning, 1995, 20(3): 273-297.

HE H, GARCIA E A. Learning from imbalanced data [J]. IEEE Transactions on Knowledge and Data Engineering, 2009, 21(9): 1263-1284.

PAN S J, YANG Q. A survey on transfer learning [J]. IEEE Transactions on Knowledge Data Engineering, 2010, 22(10): 1345-1359.

BAGHERI H, ISLAM M J. Sentiment analysis of Twitter data [EB/OL].(2017-12-16)[2018-01-05]. https:∥cn.arxiv.org/ftp/arxiv/papers/1711/1711. 10377.pdf.

YONG R, WANG C, HE X. A transfer learning based boosting model for emotion analysis [C]∥Proceedings of the IEEE International Conference on Big Knowledge. Piscataway, NJ, USA: IEEE, 2017: 264-269.

WU H, JIN Q. Improving emotion classification on Chinese microblog texts with auxiliary cross-domain data [C]∥Proceedings of the International Conference on Affective Computing and Intelligent Interaction. Piscataway, NJ, USA: IEEE, 2015: 821-826.

田锋, 兰田, CHAO Kuo-Ming, 等. 领域实例迁移的交互文本非平衡情感分类方法 [J]. 西安交通大学学报, 2015, 49(4): 67-72.

TIAN Feng, LAN Tian, CHAO Kuo-Ming, et al. A unbalanced motion classification method for interactive texts based on multiple-domain instance transfer [J]. Journal of Xi'an Jiaotong University, 2015, 49(4): 67-72.

ZHANG W, ZHANG H, WANG D, et al. Transfer learning by linking similar feature clusters for sentiment classification [C]∥Proceedings of the IEEE International Conference on Tools with Artificial Intelligence. Piscataway, NJ, USA: IEEE, 2017: 1019-1026.

LI Tao, SINDHWANI V, DING C, et al. Knowledge transformation for cross-domain sentiment classification [C]∥Proceedings of the 32nd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: ACM, 2009: 716-717.

庄福振. 迁移学习中文本分类算法研究 [D]. 北京: 中国科学院大学, 2011: 31-46.

戴昌钧. 信息效用函数与Wundt曲线 [J]. 高校应用数学学报, 1991(2): 241-252.

CHAWLA N V, BOWYER K W, HALL L O, et al. SMOTE: synthetic minority over-sampling technique [J]. Journal of Artificial Intelligence Research, 2002, 16(1): 321-357.

KIM Y. Convolutional neural networks for sentence classification [EB/OL].(2014-09-03)[2018-01-05]. https: ∥arxiv.org/abs/1408.5882.

张国和,黄凯,张斌,等.最大稳定极值区域与笔画宽度变换的自然场景文本提取方法.2017,51(1):135-140.[doi:10. 7652/xjtuxb201701021]

李扬,潘泉,杨涛.基于短文本情感分析的敏感信息识别.2016,50(9):80-84.[doi:10.7652/xjtuxb201609013]

王菲菲,杨扬,蒋飞,等.面向用户话题相似性特征的链路预测方法.2016,50(8):103-109.[doi:10.7652/xjtuxb2016 08017]

范正光,屈丹,闫红刚,等.借助音频数据的发音字典新词学习方法.2016,50(6):75-82.[doi:10.7652/xjtuxb201606012]

孙立远,周亚东,管晓宏,等.利用信息传播特性的中文网络新词发现方法.2015,49(12):59-64.[doi:10.7652/xjtuxb 201512010]

刘凯,张立民,孙永威,等.利用深度玻尔兹曼机与典型相关分析的自动图像标注算法.2015,49(6):33-38.[doi:10. 7652/xjtuxb201506006]

赵煜,邵必林,边根庆.一种融合词序信息的多粒度文本话题情感联合模型.2014,48(11):103-108.[doi:10.7652/xjtuxb 201411018]

杜友田,辛刚,郑庆华.融合异构信息的网络视频在线半监督分类方法.2013,47(7):96-101.[doi:10.7652/xjtuxb201307 018]

孙艳,周学广,付伟.无监督的主题情感混合模型研究.2013,47(4):120-125.[doi:10.7652/xjtuxb201301023]

田丰,桂小林,杨攀,等.采用类别相似度聚合的关联文本分类方法.2012,46(12):6-11.[doi:10.7652/xjtuxb201212002]

杨攀,桂小林,田丰,等.一种高效的用于话题检测的关键词元聚类方法.2012,46(10):24-28.[doi:10.7652/xjtuxb2012 10005]

霍战鹏,魏正英,张梦,等.手机短信远程控制灌溉系统.2012,46(10):36-41.[doi:10.7652/xjtuxb201210007]

豆增发,高琳.利用膜粒子群优化和信息熵的医学文本特征选择.2012,46(4):45-51.[doi:10.7652/xjtuxb201204008]

薛峰,周亚东,高峰,等.一种突发性热点话题在线发现与跟踪方法.2011,45(12):64-69.[doi:10.7652/xjtuxb201112 012]

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

An Unbalanced Emotion Classification Method for Interactive Texts Based on Multiple-Domain Instance Transfer

Turbulent Kinetic Energy Dissipation Mechanism of Supercritical Carbon Dioxide Leakage Flow in a Scallop Bionic Seal Cavity

Electrothermal Coupling Research on the Evolutionary Law of Anti-Corona Performance for High-Altitude Air-Cooled Generators

Integrated and Joint Design Optimization of Centrifugal Compressor Impeller and Casing Treatment at Multiple Working Points

Diffusion Characteristics and Influencing Factors in Case of Hydrogen Leakage from Fuel Cell Vehicles Running in Tunnels

Related Author

兰田 1

CHAO Kuo-Ming 3

高鹏达 1

Related Institution

Shaanxi Key Laboratory of Satellite and Terrestrial Network Technology Research and Development, Xi'an Jiaotong University

⁰