We propose and test a practical means of finding poor pronunciations and missing variants for large lexicons. We do so by statistically assessing the confidence of each phone in each pronunciation and comparing it with the statistical distribution of the same confidence metric for corresponding phones over the entire training corpus. A phone is targeted for correction for each word in which its...

The present study dealt with the wall static pressure of a rectangular curved duct with angle of turn 90° at all four faces with various locations. The normalized wall static pressure distributions at the top, bottom, inside and outside surfaces of the C-duct are drawn in the form of contours at three different air velocities, Uav = 20m/s, 40m/s and 60m/s respectively. These iso-bar contours in...

Author details Mayo Clinic, Rochester, MN, USA. Villa Metabolica, ZKJM, MC, University of Mainz,, Mainz, Germany. Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands. Actelion Pharmaceuticals Ltd, Allschwil, Switzerland. Numerus Ltd, Wokingham, UK. INSERM Unit 820, Lyon, France. Fundació Hospital Sant Joan de Déu, Barcelona, Spain. Department of Neurology, Mayo Clinic, ...

We extend back to 1890 the volatility implied by options index (VIX), available only since 1986, using the frequency of words on the front-page of the Wall Street Journal. News implied volatility (NVIX) captures well the disaster concerns of the average investor over this longer history. NVIX is particularly high during stock market crashes, times of policy-related uncertainty, world wars and f...

P quality is probably undervalued by firms because there is little consensus about appropriate measures and methods to research quality. We suggest that published ratings of a product’s quality are a valid source of quality information with important strategic and financial impact. We test this thesis by an event analysis of abnormal returns to stock prices of firms whose new products are evalu...

For investment decisions the development of market trends is very important. In this contribution we present our results concerning the influence of news on market trends. We processed the stock news delivered by the Wall Street Journal with two methods of text mining – Bayes classification and grammar-driven classification. We found some potentialities of Dow Jones trend prediction and present...

We evaluate the Berkeley parser on text from an online discussion forum. We evaluate the parser output with and without gold tokens and spellings (using Sparseval and Parseval), and we compile a list of problematic phenomena for this domain. The Parseval f-score for a small development set is 77.56. This increases to 80.27 when we apply a set of simple transformations to the input sentences and...

This paper proposes a new structure for use in MLLR adaptation aiming at constraining the transform for potentially better parameter estimation from sparse adaptation data. Motivations for the use of the new structure, and EM based parameter estimation are presented. Experimental results on Spoke3 of the Wall Street Journal task revealed that the proposed transformations outperform a full matri...

C often return a product to a retailer because they learn after purchase that the product does not match as well with preferences as had been expected. This is a costly issue for retailers and manufacturers— in fact, it is estimated that the U.S. electronics industry alone spent $13.8 billion dollars in 2007 to restock returned products [Lawton, C. 2008. The war on returns. Wall Street Journal ...

A data-driven approach is presented for improving the performance of separating single-channel mixed speech signals, assuming unknown, arbitrary temporal dynamics. The new approach seeks and separates the longest mixed speech segments which can be accurately matched by composite training segments. Lengthening the mixed speech segments to match reduces the uncertainty of the matching constituent...