Hierarchical Hidden Conditional Random Fields for Information Extraction
Hierarchical Hidden Conditional Random Fields for Information Extraction
复制标题
DOI：
10.1007/978-3-642-25566-3_14
复制
发表时间：
2011-01
期刊：
影响因子：
--
通讯作者：
Satoshi Kaneko;A. Hayashi;N. Suematsu;Kazunori Iwata
中科院分区：
文献类型：
--
作者：
Satoshi Kaneko;A. Hayashi;N. Suematsu;Kazunori Iwata
隐马尔可夫模型（Hidden Markov Models，HMM）是一种非常流行的时间序列数据生成模型。然而，最近的工作表明，对于许多任务，条件随机场（CRF），一种判别模型，比HALTH执行得更好。信息抽取是从文本中自动抽取指定类或关系的实例的任务。已经提出了一种使用分层隐马尔可夫模型（Hierarchical Hidden Markov Models，HHMM）的信息提取方法。HHSTOM是HSTOM的一个推广，是具有层次状态结构的生成模型。在以前的研究中，我们开发了层次隐藏条件随机场（HHCRF），一个判别模型对应HHRF。在本文中，我们提出了使用HHCRF的信息提取，然后通过实验比较HHCRF和HHCRF的性能。
Hidden Markov Models (HMMs) are very popular generative models for time series data. Recent work, however, has shown that for many tasks Conditional Random Fields (CRFs), a type of discriminative model, perform better than HMMs. Information extraction is the task of automatically extracting instances of specified classes or relations from text. A method for information extraction using Hierarchical Hidden Markov Models (HHMMs) has already been proposed. HHMMs, a generalization of HMMs, are generative models with a hierarchical state structure. In previous research, we developed the Hierarchical Hidden Conditional Random Field (HHCRF), a discriminative model corresponding to HHMMs. In this paper, we propose information extraction using HHCRFs, and then compare the performance of HHMMs and HHCRFs through an experiment.