语言建模与树替代文法Language Modeling with Tree Substitution Grammars |
|
课程网址: | http://videolectures.net/nipsworkshops09_post_lmtsg/ |
主讲教师: | Matt Post |
开课单位: | 罗切斯特大学 |
开课时间: | 2010-01-19 |
课程语种: | 英语 |
中文简介: | 我们发现,使用折叠的吉布斯采样器产生的树替换语法(TSG)比标准上下文无关语法和其他启发式训练的TSG对测试数据的困惑更低,这表明它更适合于语言建模。跨TSG派生训练一个更复杂的双线性分析模型显示出进一步的(尽管细微的)改进。我们使用TSG作为语言模型进行分析并指出未来的研究领域。 |
课程简介: | We show that a tree substitution grammar (TSG) induced with a collapsed Gibbs sampler results in lower perplexity on test data than both a standard context-free grammar and other heuristically trained TSGs, suggesting that it is better suited to language modeling. Training a more complicated bilexical parsing model across TSG derivations shows further (though nuanced) improvement. We conduct analysis and point to future areas of research using TSGs as language models. |
关 键 词: | 树替换文法; 抑癌基因; 建模语言 |
课程来源: | 视频讲座网 |
最后编审: | 2020-06-06:zyk |
阅读次数: | 23 |