Improving Parallel Corpus Quality for Chinese-Vietnamese Statistical Machine Translation

The performance of a machine translation system heavily depends on the quantity and quality of the bilingual language resource.However,getting a parallel corpus,which has a large scale and is of high quality,is a very difficult task especially for low resource languages such as Chinese-Vietnamese.Fortunately,multilingual user generated contents (UGC),such as bilingual movie subtitles,provide us access to automatic construction of the parallel corpus.Although the amount of UGC parallel corpora can be considerable,the original corpus is not suitable for statistical machine translation (SMT) systems.The corpus may contain translation errors,sentence mismatching,free translations,etc.To improve the quality of the bilingual corpus for SMT systems,three filtering methods are proposed:sentence length difference,the semantic of sentence pairs,and machine learning.Experiments are conducted on the Chinese to Vietnamese translation corpus.Experimental results demonstrate that all the three methods effectively improve the corpus quality,and the machine translation performance (BLEU score) can be improved by 1.32.     

Improving Parallel Corpus Quality for Chinese-Vietnamese Statistical Machine Translation

The performance of a machine translation system heavily depends on the quantity and quality of the bilingual language resource.However,getting a parallel corpus,which has a large scale and is of high quality,is a very difficult task especially for low resource languages such as ChineseVietnamese.Fortunately,multilingual user generated contents (UGC),such as bilingual movie subtitles,provide us access to automatic construction of the parallel corpus.Although the amount of UGC parallel corpora can be considerable,the original corpus is not suitable for statistical machine translation (SMT) systems.The corpus may contain translation errors,sentence mismatching,free translations,etc.To improve the quality of the bilingual corpus for SMT systems,three filtering methods are proposed:sentence length difference,the semantic of sentence pairs,and machine learning.Experiments are conducted on the Chinese to Vietnamese translation corpus.Experimental results demonstrate that all the three methods effectively improve the corpus quality,and the machine translation performance (BLEU score) can be improved by 1.32.     

Development of a Risk-Ranking Framework to Evaluate Potential High-Threat Microorganisms, Toxins, and Chemicals in Food

ABSTRACT:  Through a cooperative agreement with the U.S. Food and Drug Administration, the Institute of Food Technologists developed a risk-ranking framework prototype to enable comparison of microbiological and chemical hazards in foods and to assist policy makers, risk managers, risk analysts, and others in determining the relative public health impact of specific hazard–food combinations. The prototype is a bottom-up system based on assumptions that incorporate expert opinion/insight with a number of exposure and hazard-related risk criteria variables, which are propagated forward with food intake data to produce risk-ranking determinations. The prototype produces a semi-quantitative comparative assessment of food safety hazards and the impacts of hazard control measures. For a specific hazard–food combination the prototype can produce a single metric: a final risk value expressed as annual pseudo-disability adjusted life years (pDALY). The pDALY is a harmonization of the very different dose–response relationships observed for chemicals and microbes. The prototype was developed on 2 platforms, a web-based user interface and an Analytica® model (Lumina Decision Systems, Los Gatos, Calif., U.S.A.). Comprising visual basic language, the web-based platform facilitates data input and allows use concurrently from multiple locations. The Analytica model facilitates visualization of the logic flow, interrelationship of input and output variables, and calculations/algorithms comprising the prototype. A variety of sortable risk-ranking reports and summary information can be generated for hazard–food pairs, showing hazard and dose–response assumptions and data, per capita consumption by population group, and annual p-DALY.     

Metabolic mapping of the effects of oral alcohol self-administration in rats

A method for the quantitation of protein carbonyls, which have been widely employed as markers of protein oxidative damage, is described. Protein carbonyls were derivatized with tritiated sodium borohydride and the tritiated proteins were separated on SDS-PAGE. Protein bands, visualized by Coomassie blue staining, were then excised and incubated in 30% H2O2 at 60 degrees C for 48 h. Tritium, incorporated into the proteins, was quantitated by liquid scintillation counting after gel solubilization by H2O2. This method can be applied to the measurement of carbonylation of specific proteins as it employs SDS-PAGE and has the advantage that unreacted NaB3H4 in the labeling reaction mixture need not be removed. The present method, when combined with immunochemical detection of protein carbonyls, should be very useful in the quantitation of oxidative damage to individual proteins.     

An overview of recommender systems in the healthy food domain

Recently, food recommender systems have received increasing attention due to their relevance for healthy living. Most existing studies on the food domain focus on recommendations that suggest proper food items for individual users on the basis of considering their preferences or health problems. These systems also provide functionalities to keep track of nutritional consumption as well as to persuade users to change their eating behavior in positive ways. Also, group recommendation functionalities are very useful in the food domain, especially when a group of users wants to have a dinner together at home or have a birthday party in a restaurant. Such scenarios create many challenges for food recommender systems since the preferences of all group members have to be taken into account in an adequate fashion. In this paper, we present an overview of recommendation techniques for individuals and groups in the healthy food domain. In addition, we analyze the existing state-of-the-art in food recommender systems and discuss research challenges related to the development of future food recommendation technologies.     

Modelling the effects of past and future climate on the risk of bluetongue emergence in Europe

Vector-borne diseases are among those most sensitive to climate because the ecology of vectors and the development rate of pathogens within them are highly dependent on environmental conditions. Bluetongue (BT), a recently emerged arboviral disease of ruminants in Europe, is often cited as an illustration of climate''s impact on disease emergence, although no study has yet tested this association. Here, we develop a framework to quantitatively evaluate the effects of climate on BT''s emergence in Europe by integrating high-resolution climate observations and model simulations within a mechanistic model of BT transmission risk. We demonstrate that a climate-driven model explains, in both space and time, many aspects of BT''s recent emergence and spread, including the 2006 BT outbreak in northwest Europe which occurred in the year of highest projected risk since at least 1960. Furthermore, the model provides mechanistic insight into BT''s emergence, suggesting that the drivers of emergence across Europe differ between the South and the North. Driven by simulated future climate from an ensemble of 11 regional climate models, the model projects increase in the future risk of BT emergence across most of Europe with uncertainty in rate but not in trend. The framework described here is adaptable and applicable to other diseases, where the link between climate and disease transmission risk can be quantified, permitting the evaluation of scale and uncertainty in climate change''s impact on the future of such diseases.