We observe that NER label information can be used to correct alignment mistakes, and present a graphical model that performs bilingual NER tagging jointly with word alignment, by combining two monolingual tagging models with two unidirectional alignment models.

In order to model this uncertainty, we extend the two previously independent CRF models into a larger undirected graphical model , by introducing a cross-lingual edge factor gb(z', j ) for every pair of word positions (2', j) E A.

The way DD algorithms work in decomposing undirected graphical models is analogous to other message passing algorithms such as loopy belief propagation, but DD gives a stronger optimality guarantee upon convergence (Rush et al., 2010).

We introduced a graphical model that combines two HMM word aligners and two CRF NER taggers into a joint model, and presented a dual decomposition inference method for performing efficient decoding over this model.

In this work, we first develop a bilingual NER model (denoted as BI-NER) by embedding two monolingual CRF-based NER models into a larger undirected graphical model , and introduce additional edge factors based on word alignment (WA).

previous applications of the DD method in NLP, where the model typically factors over two components and agreement is to be sought between the two (Rush et al., 2010; Koo et al., 2010; DeNero and Macherey, 2011; Chieu and Teow, 2012), our method decomposes the larger graphical model into many overlapping components where each alignment edge forms a separate factor.

We introduce a cross-lingual edge factor C (i, j) in the undirected graphical model for every pair of word indices (2', j), which predicts a binary vari-

Figure 1: An example graph modeling relations between mentions.

Many graph models for coreference resolution operate on A = V x V. Our multigraph model allows us to have multiple edges with different labels between mentions.

In contrast to previous work on similar graph models we do not learn any edge weights from training data.

Our approach belongs to a class of recently proposed graph models for coreference resolution (Cai and Strube, 2010;

The graph model described in Section 3 is based on expressing relations between pairs of mentions via edges built from such relations.

Figure 1 shows the graphical model for our joint Bigram model (the Unigram case is trivially recovered by generating the Ums directly from L rather than from LUi,j_1).

Figure 2 gives the mathematical description of the graphical model and Table 1 provides a key to the variables of our model.

Figure l: The graphical model for our joint model of word-final /t/-deletion and Bigram word segmentation.