SVM-based Interactive Document Retrieval with Active Learning

SVM-based Interactive Document Retrieval with Active Learning
复制标题

基于 SVM 的主动学习交互式文档检索

DOI:
10.1007/s00354-007-0034-4
复制
发表时间:
2007
影响因子:
2.6
通讯作者:
S. Yamada
S. Yamada
中科院分区:
计算机科学4区
文献类型:
--
作者:
T. Onoda;H. Murata;S. Yamada

文献摘要

参考文献

被引文献

相似文献

本文描述了一种基于主动学习的支持向量机(SVM)在交互式文档检索中的应用。将SVM等分类学习应用到相关反馈中,已经取得了一些成功的成果。但在文献检索中没有充分利用样本分布的特点。根据文献检索中样本的分布情况,提出了启发式方法来对文献显示进行偏置,以供用户判断。这种启发式算法通过选择实例在正支持向量的邻域中显示用户来执行,提高了学习效率。我们使用我们提出的启发式实现了一个基于svm的交互式文档检索系统,并将其与基于rocchio的传统系统和不使用启发式的基于svm的系统进行了比较。我们使用超过50万篇报纸文章的大数据集进行了系统实验,并证实了我们的系统优于其他系统。
This paper describes an application of SVM (Support Vector Machines) to interactive document retrieval using active learning. Some works have been done to apply classification learning like SVM to relevance feedback and have obtained successful results. However they did not fully utilize characteristic of example distribution in document retrieval. We propose heuristics to bias document showing for user’s judgement according to distribution of examples in document retrieval. This heuristics is executed by selecting examples to show a user in neighbors of positive support vectors, and it improves learning efficiency. We implemented a SVM-based interactive document retrieval system using our proposed heuristics, and compared it with conventional systems like Rocchio-based system and a SVM-based system without the heuristics. We conducted systematic experiments using large data sets including over 500,000 newspaper articles and confirmed our system outperformed other ones.
DOI: --
发表时间: 2001-06
期刊: --
影响因子: --
作者:
H. Drucker;B. Shahraray;D. Gibbon
通讯作者: H. Drucker;B. Shahraray;D. Gibbon
DOI: 10.1109/ijcnn.2006.246829
发表时间: 2006-10
期刊: The 2006 IEEE International Joint Conference on Neural Network Proceedings
影响因子: --
作者:
T. Onoda;H. Murata;S. Yamada
通讯作者: T. Onoda;H. Murata;S. Yamada