Exploiting Chinese character models to improve speech recognition performance

The Chinese language is based on characters which are syllabic in nature. Since languages have syllabotactic rules which govern the construction of syllables and their allowed sequences, Chinese character sequence models can be used as a first level approximation of allowed syllable sequences. N-gram character sequence models were trained on 4.3 billion characters. Characters are used as a first level recognition unit with multiple pronunciations per character. For comparison the CU-HTK Mandarin word based system was used to recognize words which were then converted to character sequences. The character only system error rates for one best recognition were slightly worse than word based character recognition. However combining the two systems using log-linear combination gives better results than either system separately. An equally weighted combination gave consistent CER gains of 0.1-0.2% absolute over the word based standard system. Copyright © 2009 ISCA.

Structure learning in a sensorimotor association task.

Learning is often understood as an organism's gradual acquisition of the association between a given sensory stimulus and the correct motor response. Mathematically, this corresponds to regressing a mapping between the set of observations and the set of actions. Recently, however, it has been shown both in cognitive and motor neuroscience that humans are not only able to learn particular stimulus-response mappings, but are also able to extract abstract structural invariants that facilitate generalization to novel tasks. Here we show how such structure learning can enhance facilitation in a sensorimotor association task performed by human subjects. Using regression and reinforcement learning models we show that the observed facilitation cannot be explained by these basic models of learning stimulus-response associations. We show, however, that the observed data can be explained by a hierarchical Bayesian model that performs structure learning. In line with previous results from cognitive tasks, this suggests that hierarchical Bayesian inference might provide a common framework to explain both the learning of specific stimulus-response associations and the learning of abstract structures that are shared by different task environments.

Beyond words: Testing alignment among inter-organizational performance measures

Purpose: This paper seeks to measure in a quantitative way the degree of alignment among a set of performance measures between two organizations. Design/methodology/approach: This paper extends Venkatraman's test of coalignment to assess the alignment of a set of performance measures governing a contractual inter-organizational relationship. The authors applied the test and present coefficients of misalignment across three sets of measures: those used by a service provider involved in the research, those used by customers contracting the services, and those documented in 11 contracts studied. Findings: Results confirmed a high degree of alignment between target and actual operational performance in the contracts. The alignment of customers' financial objectives and contracts' operational metrics was low. Calculations show poor alignment between the objectives of the provider and the contribution received from the contracts. Research limitations/implications: Some limitations of the conclusions include the small sample of contracts used in the calculations. Further research should include not only actual contracts, but also failed ones. Practical implications: It is possible that misaligned goals, represented in misaligned performance measures, lead to tensions in intra-firm relationships. If these tensions are not addressed properly the relationship could be unstable or terminated prematurely. This method of measuring alignment could detect early potential dangers in intra-firm relationships. Originality/value: This paper extends Venkatraman's test of coalignment to assess the alignment of a set of performance measures governing a contractual inter-organizational relationship. Management researchers and business professionals may use this methodology when exploring degrees of alignment of performance measures in intra-functional and inter-firm relationships. © Emerald Group Publishing Limited.

Graphone model interpolation and Arabic pronunciation generation

This paper extends n-gram graphone model pronunciation generation to use a mixture of such models. This technique is useful when pronunciation data is for a specific variant (or set of variants) of a language, such as for a dialect, and only a small amount of pronunciation dictionary training data for that specific variant is available. The performance of the interpolated n-gram graphone model is evaluated on Arabic phonetic pronunciation generation for words that can't be handled by the Buckwalter Morphological Analyser. The pronunciations produced are also used to train an Arabic broadcast audio speech recognition system. In both cases the interpolated graphone model leads to improved performance. Copyright © 2011 ISCA.