We present a novel class of learning algorithms for undirected graphical models, based on the contrastive free energy (CF). In particular we study the naive mean field, TAP and Bethe approximations to the contrastive free energy. The main advantage of CFlearning is the fact that it eliminates the need to infer equilibrium statistics for which mean field type approximations are particularly unsu...

Narrowly defined, contrastive linguistics can be regarded as a branch of comparative linguistics that is concerned with pairs of languages which are ‘socio-culturally linked’. Two languages can be said to be socio-culturally linked when (i) they are used by a considerable number of bior multilingual speakers, and/or (ii) a substantial amount of ‘linguistic output’ (text, oral discourse) is tran...

In this study we show that constituency is of limited importance for a proper treatment of the interaction between the linear position of a wa-marked nominal in a Japanese sentence and possible domains of contrastive focus, and that constraints concerning contrastive focus should be represented in terms of linear order and not constituency. Linearisation HPSG, where linear order is independent ...

This study investigates how Seoul Korean-speaking children and adults use pitchand duration-related phonetic cues to mark focus. It was found that to distinguish focus from non-focus, the adults used both the pitchand duration-related cues, but the children used only the duration-related cues to distinguish focus from post-focus. Further, neither the adults nor the children distinguish narrow f...

We propose a neural probabilistic structured-prediction method for transition-based natural language processing, which integrates beam search and contrastive learning. The method uses a global optimization model, which can leverage arbitrary features over nonlocal context. Beam search is used for efficient heuristic decoding, and contrastive learning is performed for adjusting the model accordi...

The present article describes the development of L1-L2map, a multi-lingual contrastive analysis tool. It uses the phoneme inventories of a large number of languages, but also contains more detailed phonetic information. An example of this is the information about the syllable positions in which the sounds can occur in a given language, which is very useful for computer-assisted pronunciation tr...

We generalize contrastive estimation in two ways that permit adding more knowledge to unsupervised learning. The first allows the modeler to specify not only the set of corrupted inputs for each observation, but also how bad each one is. The second allows specifying structural preferences on the latent variable used to explain the observations. They require setting additional hyperparameters, w...

Being able to predict the placement of con-trastive accent is essential for the assignment of correct accentuation patterns in spoken language generation. I discuss two approaches to the generation of contrastive accent and propose an alternative method that is feasible and computationally attractive in data-to-speech systems. 1 Motivation The placement of pitch accent plays an important role i...