TL;DR本文介绍了一种 CRFs 的特征归纳方法,该方法基于构建仅显著增加对数似然的特征联合的原则,与 Della Pietra 等人的方法不同,具有更高的准确性和较低的特征数量,适用于线性链 CRFs 和关系马尔可夫网络,实验证明其在命名实体识别任务中的有效性。
Abstract
conditional random fields (CRFs) are undirected graphical models, a special
case of which correspond to conditionally-trained finite state machines. A key
advantage of these models is their great flexibility to include a wide array of
overlapping, multi-granularity, non-independent fea