共查询到12条相似文献,搜索用时 46 毫秒
1.
利用术语相似度将同义词间的相似程度数量化,以此量化关系对用于信息检索的简单贝叶斯网络进行若干改进,构造一个四层贝叶斯网络检索模型。给出新模型的拓扑结构、各层节点详尽的概率估计以及文档检索与推理过程。最后,对新模型进行评估,结果表明该模型可以有效地提高检索性能,在一定程度上实现基于语义的信息检索,这正是目前信息检索发展的必然趋势。 相似文献
2.
3.
4.
个性化信息服务越来越成为信息检索领域研究的热点.将贝叶斯网络和互信息相结合,用于个性化检索的用户建模中,建立了一个能同时表达特征词的统计分布和特征词间的语义相关性的用户模型.模型以贝叶斯网络结构为框架,包含了特征词的概率统计信息和特征词间互信息,并引入了时间机制.实验结果表明,用该模型进行信息检索,在查全率和查准率方面都得到了提高. 相似文献
5.
6.
针对传统的信息检索方法无法实现用户查询的语义理解、检索效率低等问题,本文提出基于领域本体进行查询扩展的贝叶斯网络检索模型。该模型首先将用户查询通过领域本体进行语义扩展,然后将扩展后的查询作为证据在贝叶斯网络检索模型中进行传播,进而得到查询结果,实验表明本文提出的贝叶斯网络检索模型能提高检索效率。 相似文献
7.
8.
9.
研究表明合理考虑术语之间的关系可以提高检索系统的性能。采用共现分析的方法从文档集合中学习得到术语之间的关系,并应用到结构化文档检索中,提出了一个基于贝叶斯网络的结构化文档检索模型,给出了其拓扑结构、概率估计以及推理过程。实验表明该模型的检索性能要优于没有考虑术语之间关系的模型。 相似文献
10.
11.
Bayesian network based business information retrieval model 总被引:3,自引:3,他引:0
The quality of business information can significantly affect the operation level of enterprise. This paper analyses the problem
of business information retrieval (BIR). A Bayesian Network Based business information retrieval model (BN-BIRM) is proposed
by means of Bayesian network (BN) and information retrieval (IR) theory and a method for query adaptation is presented. In
this model the customized query requirement of enterprise (CQR) is expressed in terms of the predefined illustrative documents
related to business domain. The similarities between the documents and the query are evaluated with the conditional probabilities
among the nodes in the BN. In the experiments, BN-BIRM is compared with the Belief Network model based on vector space model
(VSM) ranking strategy and the Inference Network model based on TF-IDF ranking strategy. The experimental results show that
BN-BIRM is effective for collecting business information on a large scale.
相似文献
Zheng WangEmail: |
12.
Boosted Bayesian network classifiers 总被引:2,自引:0,他引:2
The use of Bayesian networks for classification problems has received a significant amount of recent attention. Although computationally
efficient, the standard maximum likelihood learning method tends to be suboptimal due to the mismatch between its optimization
criteria (data likelihood) and the actual goal of classification (label prediction accuracy). Recent approaches to optimizing
classification performance during parameter or structure learning show promise, but lack the favorable computational properties
of maximum likelihood learning. In this paper we present boosted Bayesian network classifiers, a framework to combine discriminative
data-weighting with generative training of intermediate models. We show that boosted Bayesian network classifiers encompass
the basic generative models in isolation, but improve their classification performance when the model structure is suboptimal.
We also demonstrate that structure learning is beneficial in the construction of boosted Bayesian network classifiers. On
a large suite of benchmark data-sets, this approach outperforms generative graphical models such as naive Bayes and TAN in
classification accuracy. Boosted Bayesian network classifiers have comparable or better performance in comparison to other
discriminatively trained graphical models including ELR and BNC. Furthermore, boosted Bayesian networks require significantly
less training time than the ELR and BNC algorithms. 相似文献