NTCIR-3 CLIR Experiments at MSRA
NTCIR-3 CLIR Experiments at MSRA
复制标题
MSRA 的 NTCIR-3 CLIR 实验
DOI：
--
复制
发表时间：
2002
期刊：
影响因子：
--
通讯作者：
Jianfeng Gao
中科院分区：
文献类型：
--
作者：
Hongzhao He;Jianfeng Gao
本文介绍了三种统计模型，以解决跨语言信息检索（CLIR）的查询翻译歧义。首先，提出了一个衰减共生模型。它是传统同现模型的扩展，因为它包含一个衰减因子，当项之间的距离增加时，该衰减因子会降低互信息。其次，描述了一个短语翻译模型，旨在检测和翻译未存储在词典中的名词短语。最后，提出了一个三重翻译模型，提供了一种利用语言依赖信息的方法。我们使用这些模型TREC和NTCIR语料库的实验改进。
This paper describes three statistical models for the purpose of resolving query translation ambiguity for cross-language information retrieval (CLIR). First, a decaying co-occurrence model is present. It is an extension of traditional co-occurrence models in that it contains a decaying factor which decreases the mutual information when the distance between the terms increases. Second, a phrase translation model is described aiming to detect and translate noun phrases that are not stored in the dictionary. Finally, a triple translation model is proposed which provides a way of exploiting linguistic dependency information. We show experimentally improvements of using these models on TREC and NTCIR corpus.