Maximum entropy Markov models for information extraction and segmentation

Introduction

… • For each s’ separately conditional probabilities P(s|s’,o) are given by an exponential model
• Each exponential model is trained via maximum entropy … Apply Generalized IterativeScaling (GIS) for each s’ using its〈o,s)set to learn the maximum entropy solution for the transition function of s’. … – Empirical probability that the actual and the predicted segmentation agree on the placement of two lines according to some distance distribution D between lines…