E-portfolios are considered an effective way of motivating learners to review language and help reflect on their own objectives of learning and achievement. This paper describes and discusses core aspects of the e-portfolio assessment instrument used to evaluate the students’ learning in an English for Specific Purposes (ESP) course at a Thai university. It is the writer’s motivation to explore e-portfolio assessment to try out an optional formative assessment instrument which could provide guidance on students’ progress and check how far students have mastered what they have learned in their English course each semester. This way, assessment of learning can be used both to measure and promote learning. The instrument, in its pilot stage, if successfully implemented, can be optimized in other language courses offered at university.

Cyclic loading of metallic structures such as bridges leads to fatigue cracks, and in turn when the cracks propagate, the structural system accumulates damage thereby leading to serviceability loss and/or eventual collapse. These failures can be prevented by appropriate maintenance scheduling and repair. A robust maintenance strategy for metallic structures under fatigue is presented, which works with reliability metrics redefined within the framework of imprecise probabilities. The effects of uncertainties are expressed in terms of probabilities of failure, repair, and expected costs of operation. A welded connection between a web stiffener and the flange of a bridge girder is presented to illustrate and to discuss the suggested approach as well as its applicability. The proposed robust maintenance strategy is implemented in OpenCossan; the open source engine of COSSAN software for uncertainty quantification and risk management. The initial crack length is modelled as lognormal variable with imprecision of the mean value and defined standard deviation: LN ([0.5, 3], 0.4) using Reliability- Based Optimization method.

Background: The COVID-19 pandemic created challenges for countries and the health system. Nurses face a strange challenge. On the one hand, they want to do their job properly, and on the other hand, personal protective equipment (PPE) is not sufficiently available. \nObjective: To show how Iranian nurses are faced to COVID-19 disaster.\nDesign: A qualitative content analysis study\nSetting: A university hospital in Iran, Tehran\nParticipants: 16 participants including 8 nurses, 5 supervisor and 3 head nurse.\nMethods: Using from purposive sampling, nurses who were working with COVID-19 patients were interviewed during February to March 2020. Data was analyzed using the MAXQDA 10 software and code were categorized based on similarities and differences. \nFinding: Data analysis revealed 212 codes, then they were classified based on similarities and differences in to the 16 categories and then four themes including: \"unpreparedness\", \"positive adaptation\", \"negative coping\", and \"reorganization”.\nConclusion: Given that nurses are at the frontline of the fighting, it is a good opportunity to demonstrate the role of nurses in reducing the burden of illness, as well as identifying problems and opportunities and planning appropriate interventions.

This study aimed at examining the causal effects of linkage models for rice production and distribution on farmers’ outcome perforamces, such as farm income, yield and output price. The study employed a propensity score matching approach to measure the impact of farmers’ participation in the linkage models on the crop outcomes by using a cross-sectional data gathered from 360 rice growers in Vietnam’s Mekong Delta. The findings of the study revealed that participation in the linkage models had positive impacts on rice incomes and output price. This paper significantly contributes to fill a considerable gap in effect estimation of rice production and distribution under a contract.

We investigated physical, chemical and biogeochemical parameters in the surface waters of three coastal areas characterized by different degrees of pollution (the Sfax Northern (NC) and Southern coasts (SC) and in the Ghannouch area of Gabes (GA)) during October and November 2014 in order to identify pertinent tracers of anthropogenic inputs. The SIAPE (SC) and the GCT (GA) were the main sources of inorganic pollutants such as PO43-, HCO3-, Cl- and SO42-. High contents in total polyphenolic and flavonoid compounds were observed in the SC due to the olive oil treatment waste water released from margins contributing to the increase of COD. The BOD5/COD ratio was < 0.5 for 50% of stations; it thus underscored a pollution of chemical origin. The COD/BOD5 ratio was > 3 just for stations 2, 4 and 5 (SC); which underlined an organic load resistant to biodegradation due to the high content in polyphenolic compounds.

In present study, a method to increase the Gurmukhi(Indian Regional Language) Handwritten Digit recognition accuracy on small datasets has been proposed. The proposed methodology uses a cascaded DCNN (Deep Convolutional Neural Network) with XGBoost(Extreme Gradient Boosting) algorithm. Also a comprehensive analysis has been done to apprehend the impact of kernel size of DCNN on recognition accuracy. The reason for using DCNN is its impressive recognition accuracy of classification for handwritten digits, but in order to achieve a better recognition accuracy DCNN requires a huge amount of data and also significant training/testing time. In order to increase the accuracy of DCNN for a small dataset more images have been generated by applying shear transformation (A transformation which preserves parallelism but not length and angles) to the original images. To address the issue of large training time only two hidden layers along with selective cascading XGBoost among the misclassified digits have been used. The proposed methodology have led to increase in overall rate of recognition with a marginal increase in testing time. Also the issue of overfitting is discussed in detail and has been reduced to a great extent. Finally the results were compared with performance of SVM(Support Vector Machine) classifier on DCT(Discrete Cosine Transform) and DWT(Discrete Wavelet Transform) Features obtained on the same dataset.