We hypothesise that general numerical optimisation techniques result in improved performance over iterative scaling algorithms for training CRFs. In Structural, Syntactic, and Statistical Pattern Recognition; Lecture Notes in Computer Science, Vol. These methods include sliding window methods, recurrent sliding windows, hidden Markov models, conditional random fields, and graph transformer networks. In Proceedings of the 2003 Human Language Technology Conference and North American Chapter of the Association for Computational Linguistics (HLT/NAACL-03), 2003.Experiments run on a subset of a well-known text chunking data set confirm that this is indeed the case. The paper also discusses some open research issues. Conditional random fields for sequence labeling offer advantages over both generative models like HMMs and classifers applied at each sequence position.Additionally, CRFs avoid the label bias problem, a weakness exhibited by maximum entropy Markov models (MEMMs) and other conditional Markov models based on directed graphical models. We present conditional random fields, a framework for building probabilistic models to segment and label sequence data.
The underlying idea is that of defining a conditional probability distribution over label sequences given a particular observation sequence, rather than a joint distribution over both label and observation sequences. Department of Computer and Information Science, University of Pennsylvania, 2004. Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data.
The primary advantage of CRFs over hidden Markov models is their conditional nature, resulting in the relaxation of the independence assumptions required by HMMs in order to ensure tractable inference. In Proceedings of the Eighteenth International Conference on Machine Learning (ICML-2001), 2001.