Phrase-based statistical machine translation: explanation of its processes and statistical models and evaluation of the English to Spanish translations produced

Statistical machine translation (SMT) is an approach to Machine Translation (MT) that uses statistical models whose parameter estimation is based on the analysis of existing human translations (contained in bilingual corpora). From a translation student’s standpoint, this dissertation aims to explain how a phrase-based SMT system works, to determine the role of the statistical models it uses in the translation process and to assess the quality of the translations provided that system is trained with in-domain goodquality corpora. To that end, a phrase-based SMT system based on Moses has been trained and subsequently used for the English to Spanish translation of two texts related in topic to the training data. Finally, the quality of this output texts produced by the system has been assessed through a quantitative evaluation carried out with three different automatic evaluation measures and a qualitative evaluation based on the Multidimensional Quality Metrics (MQM).

No anticorrelation between cyclotron line energy and X-ray flux in 4U 0115+634

We report on an outburst of the high mass X-ray binary 4U 0115+634 with a pulse period of 3.6 s in 2008 March/April as observed with RXTE and INTEGRAL. During the outburst the neutron star’s luminosity varied by a factor of 10 in the 3–50 keV band. In agreement with earlier work we find evidence of five cyclotron resonance scattering features at ~10.7, 21.8, 35.5, 46.7, and 59.7 keV. Previous work had found an anticorrelation between the fundamental cyclotron line energy and the X-ray flux. We show that this apparent anticorrelation is probably due to the unphysical interplay of parameters of the cyclotron line with the continuum models used previously, e.g., the negative and positive exponent power law (NPEX). For this model, we show that cyclotron line modeling erroneously leads to describing part of the exponential cutoff and the continuum variability, and not the cyclotron lines. When the X-ray continuum is modeled with a simple exponentially cutoff power law modified by a Gaussian emission feature around 10 keV, the correlation between the line energy and the flux vanishes, and the line parameters remain virtually constant over the outburst. We therefore conclude that the previously reported anticorrelation is an artifact of the assumptions adopted in the modeling of the continuum.