A Conditional Random Field for Discriminatively-trained Finite-state String Edit Distance
This work addresses sequence similarity measurement for applications in data mining and bioinformatics, presenting an incremental improvement over generative models like pair HMMs.
The paper tackles the problem of measuring sequence similarity across domains like information extraction and biological analysis by introducing discriminative string-edit CRFs, a finite-state conditional random field model that enables the use of complex features and training on both positive and negative string pairs, with positive experimental results reported on several datasets.
The need to measure sequence similarity arises in information extraction, object identity, data mining, biological sequence analysis, and other domains. This paper presents discriminative string-edit CRFs, a finitestate conditional random field model for edit sequences between strings. Conditional random fields have advantages over generative approaches to this problem, such as pair HMMs or the work of Ristad and Yianilos, because as conditionally-trained methods, they enable the use of complex, arbitrary actions and features of the input strings. As in generative models, the training data does not have to specify the edit sequences between the given string pairs. Unlike generative models, however, our model is trained on both positive and negative instances of string pairs. We present positive experimental results on several data sets.