13,714 research outputs found
ITC-irst at the 2006 TC-STAR SLT Evaluation Campaign
No full textThis paper reports on the participation of ITC-irst in the 2006 Spoken Language Translation Evaluation Campaign organized by the
TC-STAR project. ITC-irst submitted runs for all translation directions, namely Spanish-to-English, English-to-Spanish and Chinese-to-
English, and types of input, that is final text edition, human verbatim transcriptions and speech recognition output. Official results show
that translations produced by our systems rank among the best ones. With respect to the translation systems we developed for the 2005
evaluation, BLEU scores were improved in every condition, from 17% up to 40% relative
TC-Release++: an efficient timestamp-based coherence protocol for many-core architectures
No full textAs we enter the era of many-core, providing the shared memory abstraction through cache coherence has become progressively difficult. The standard directory-based coherence does not scale well with increasing core count. Timestamp-based hardware coherence protocols introduced recently offer an attractive alternative solution. This paper proposes a timestamp-based coherence protocol, called TC-Release++, that efficiently supports cache coherence in large-scale systems. Our approach is inspired by TC-Weak, a recently proposed timestamp-based coherence protocol targeting GPU architectures. We first design TC-Release in an attempt to straightforwardly port TC-Weak to general-purpose many-cores. But re-purposing TC-Weak for general-purpose many-core architectures is challenging due to significant differences both in architecture and the programming model. Indeed the performance of TC-Release turns out to be worse than conventional directory protocols. We overcome the limitations and overheads of TC-Release by exploiting simple hardware support to eliminate frequent memory stalls, and an optimized lifetime prediction mechanism to improve cache performance. The resulting optimized coherence protocol TC-Release++ is highly scalable (storage scales logarithmically with core count) and shows better performance (3.0%) and comparable network traffic (within 1.3%) relative to the baseline MESI directory protocol. We use Murphi to formally verify that TC-Release++ is error-free and imposes small verification cost
Stabilization of optical vortices in noninstantaneous self-focusing medium by small rotating intensity modulation
No full textIdentification of clinically relevant viridans group streptococci by sequence analysis of the 16S-23S ribosomal DNA spacer region. J Clin Microbiol
No full textEvaluation of screening for nasopharyngeal carcinoma: trial design using Markov chain models
No full text
Characterization of the ultrathin HfO2 and Hf-silicate films grown by atomic layer deposition
No full textMechanisms for accumulation and migration of technetium-99 in saltmarsh sediments
No full textThis thesis describes the development of analytical methods for both the bulk determination of 99Tc, and determination of 99Tc in sequential extracts from sediments. These methods have been used to collect data, which, along with trace and major element data have been used to interpret the mechanisms for 99Tc input, migration and accumulation in saltmarshes. The inventory of 99Tc stored in the Thornflatt Saltmarsh, Esk Estuary has also been determined. The routine determination of 99Tc in bulk samples uses 99mTc as a yield monitor. Samples are ignited stepwise to 550°C and the 99Tc is extracted using 8M nitric acid. Many contaminants are precipitated with Fe(OH)3 and the Tc in the supernant is pre-concentrated and further purified using anion-exchange chromatography. Final separation of Tc from Ru is achieved by extraction of Tc into 5% TnOA in xylene from 2M sulphuric acid. The yield is determined by γ-spectrometric analysis of 99mTc. Determination of 99Tc is made by liquid scintillation counting. Typical recoveries are in the order of 70-95% and the method has a detection limit of 1.7 Bq/kg for a sample size of 10g. Determination of Tc in sequential extracts uses operationally defined procedures to extract: exchangeable Tc, reducible Tc and oxidisable Tc. An initial water wash is used to extract any occluded Tc and a final leach in 8 M nitric acid is used to dissolve any residual Tc. The isolation of 99Tc uses TEVA resin for Extracts 1-4 and the decontamination procedure developed for bulk analysis for Extract 5. 99mTc was used as a yield monitor, and determination of 99Tc is by liquid scintillation counting. Limits of detection were dependent on the amount of 99mTc tracer used but were found to be as low as 2.4 Bq/kg for a sample size of 2g. A study was made of the mechanisms responsible for the accumulation and migration of Tc in estuarine sediments using sediments collected from saltmarshes at Thornflatt, Carlaverock and the Ribble Estuary. 99Tc was present at determinable activities in all the sediment cores taken from these sites. Good correlations between Tc and CaO as well as CO3 concentrations and poor correlation between Tc and radionuclides adsorbed to inorganic detritus infer a direct input of 99Tc to marsh sediments. Determination of 99Tc in biota living on the marsh also showed that this was not a significant pathway for input of Tc to the sediments. Sequential extraction data imply sorption to an organic fraction of the sediment. Stable element and sequential extraction data indicates that Tc is readily oxidised and remobilised before reprecipitation where redox conditions are favourable. Data indicate a reduction potential between those of the MnIV to MnII reaction and the FeIII to FeII reaction is necessary for re-accumulation to occur, as suggested by published thermodynamic data. Data collected from reducing sediments imply that similar mechanisms are responsible for the accumulation of Mn (e.g. reduction by sulphate reducing bacteria) and the accumulation of Tc. The inventory of 99Tc held within the Thornflatt saltmarsh is proportionally less than that of 137Cs or 241Am when compared to discharges from Sellafield. However a higher proportion of 99Tc is transferred from Seliafield and incorporated into saltmarsh sediments than is suggested by previously published standard distribution coefficient data. Saltmarsh sediments are therefore a more important sink of 99Tc than extrapolations made from inventories of other radionuclides would suggest
Rough set theory with discriminant analysis in analyzing electricity loads
No full text[[abstract]]With the ability to deal with both numeric and nominal information, rough set theory (RST), which can express knowledge in a rule-based form, has been one of the most important techniques in data analysis. However, applications of rough set theory for analyzing electricity loads are not widely discussed. Thus. this investigation employs rough set theory to analyze electricity loads. Additionally, to reduce the time generating reducts by rough set theory, linear discriminant analysis (LDA) is used to generate a reduct for rough set model. Therefore, this study designs a hybrid discriminant analysis and rough set model (DARST) to provide decision rules representing relations in an electric load information system. In this investigation, nine condition factors and variations of electricity loads are employed to examine the feasibility of the hybrid model. Experimental results reveal that the proposed model can efficiently and accurately analyze the relation between condition variables and variations of electricity loads. Consequently, the proposed model is a promising alternative for developing an electric load information system and offers decision rules base for the utility management as well as operations staff. (c) 2008 Elsevier Ltd. All rights reserved.[[note]]SC
- …
