Machine learning techniques have long been the foundations of speech processing. Bayesian classification, decision trees, unsupervised clustering, the EM algorithm, maximum entropy, etc. are all part of existing speech recognition systems. Meanwhile, the success of statistical speech recognition has led to the rise of statistical and empirical methods in natural language processing.

Many of the machine learning techniques in language processing, from statistical part-of-speech tagging to the noisy channel model for machine translation have roots in work conducted in the speech field. In turn, advances in Learning Theory and algorithmic Machine Learning approaches have also made a mark on natural language and speech processing. Approaches such as memory based learning, a range of linear classifiers such as Boosting, SVMs and SNoW and others have been successfully applied to a broad range of natural language problems, and these now inspire new research in speech retrieval and recognition. We have seen an increasingly close collaboration between voice and language processing researchers in some of the shared tasks such as spontaneous speech recognition and understanding, voice data information extraction, and machine translation.

The purpose of this special issue is to invite speech and language researchers to communicate with each other, and with the machine learning community on the latest machine learning advances in their work. We hope to promote both the development of new theoretical frameworks and of further application of machine learning techniques in new ways to both speech and language areas, fueling the synergy between the two.

Papers are invited on learning applied to all speech and natural language tasks including, but not limited to:

Acoustics & Phonetics, Syntax, Semantics, Discourse and Dialog, Language Modeling, Spoken Language Understanding and Generation, Multilingual Processing, Machine Translation, Spoken Language Information Extraction and Retrieval, Natural Language and Spoken Language based Interactive Systems.

We welcome work within any machine learning and statistical frameworks and/or the development of a new framework for any of the above areas.

Original theoretical or experimental papers showing significant contribution in the above areas are invited. Papers showing the synergy between speech and language processing using learning are especially encouraged. Papers will be evaluated by experts in the relevant area of natural language learning, but should be written to be reasonably accessible to a general machine learning audience.