Search

A rapid and accurate quantitative method of high performance liquid chromatography (HPLC) with fluorescence detector has been developed for the analysis of 18 kinds of amino acids in fresh tea leaves. The samples were minced and mixed, and extracted with ultra pure water at 90℃ for 20 min. The 6-aminoquinolylN-hydroxy-succinimidyl carbamate (AQC) was used as pre-column derivatization reagent. Gradient HPLC separation was performed on a C18 column (Symmetry C18, 3.9 mm × 15 cm, 4 μm). Good linearity between concentrations and peak areas was achieved in the concentration range of 5.0-250 μmol/L for 18 kinds of amino acids. The method was validated by the analysis of five replicates. The 18 kinds of amino acid standards were spiked in fresh tea leaf samples and the average recoveries were 86.25%-109.05% with relative standard deviations (n = 5) ranging from 6.03% to 10.56%. The limit of detection (LOD) for the analytes was 0.05-1.27 μmol/L. The method was successfully applied to the analysis of the 18 kinds of amino acids in fresh tea leaves from east Dongting and west Dongting mountains in Suzhou. The results indicate that the method is simple, rapid, precise and reliable.

The interest on emotional computing has been increasing as many applications were in demand by multiple markets. This paper mainly focuses on different learning methods, and has implemented several methods: Support Vector Machine (SVM) and Deep Boltzmann Machine (DBM) for facial emotion recognition. The training and testing data sets of facial emotion prediction are from FERA 2015, and geometric features and appearance features are combined together. Different prediction systems are developed and the prediction results are compared. This paper aims to design an suitable system for facial emotion recognition.

This paper presents a novel reranking model, future reward reranking, to re-score the actions in a transition-based parser by using a global scorer. Different to conventional reranking parsing, the model searches for the best dependency tree in all feasible trees constraining by a sequence of actions to get the future reward of the sequence. The scorer is based on a first-order graph-based parser with bidirectional LSTM, which catches different parsing view compared with the transition-based parser. Besides, since context enhancement has shown substantial improvement in the arc-stand transition-based parsing over the parsing accuracy, we implement context enhancement on an arc-eager transition-base parser with stack LSTMs, the dynamic oracle and dropout supporting and achieve further improvement. With the global scorer and context enhancement, the results show that UAS of the parser increases as much as 1.20% for English and 1.66% for Chinese, and LAS increases as much as 1.32% for English and 1.63% for Chinese. Moreover, we get state-of-the-art LASs, achieving 87.58% for Chinese and 93.37% for English.

Voice conversion aims to modify the characteristics of one speaker to make it sound like spoken by another speaker without changing the language content. This task has attracted considerable attention and various approaches have been proposed since two decades ago. The evaluation of voice conversion approaches, usually through time-intensive subject listening tests, requires a huge amount of human labor. This paper proposes an automatic voice conversion evaluation strategy based on perceptual background noise distortion and speaker similarity. Experimental results show that our automatic evaluation results match the subjective listening results quite well. We further use our strategy to select best converted samples from multiple voice conversion systems and our submission achieves promising results in the voice conversion challenge (VCC2016). Published version

Large-capacity wind power will have impact on transmission line protection and auto-reclosing when short circuit occurs. This paper analyses the characteristics of wind farm model and the influences of wind power to transmission line differential protection as well as its auto-reclosing, then proposed countermeasures to meet and overcome the problem.

In this study, we outline a potential problem in normalising texts that are based on a modified version of the Arabic alphabet. One of the main resources available for processing resource-scarce languages is raw text collected from the Internet. Many less-resourced languages, such as Kurdish, Farsi, Urdu, Pashtu, etc., use a modified version of the Arabic writing system. Many characters in harvested data from the Internet may have exactly the same form but encoded with different Unicode values (ambiguous characters). The existence of ambiguous characters in words leads to word duplication, thus it is important to identify and unify ambiguous characters during the normalisation stage. Here, we demonstrate cases related to ambiguous Kurdish and Farsi characters and propose a semi-automatic approach to identifying and unifying them.

Verb forms a major category of lexicon in any language. As our daily actions consist of mostly hand actions, it is important for bilingual speakers to be able to use different specific hand action verbs to describe the corresponding actions. This study focuses on one particular type of hand action, namely, holding actions, to examine Singaporean bilingual preschoolers’ competencies in both English and Mandarin through their usage of holding verbs. Thirty bilingual children between the ages of 3 and 6 were recruited for the study. In the experiment of the study, we used the standard PPVT-IV pictures and self-selected pictures of different holding actions as stimulus and asked the children to describe the actions in English and Chinese. The results show that most of the children used the Mandarin word “ná” (拿) and the English word “hold” for almost all the scenarios of holding in the experiment. MOE (Min. of Education, S’pore) Accepted version

International audience

The Chinese verb “dǎ” is a polysemous and frequently used verb. Studies have shown that it is one of the earliest verbs acquired by monolingual children by the age of five year old, they can use most of the commonly used senses in their daily life. But whether it is an easy task for bilingual children to acquire and use the verb in different contexts is unknown. Our study investigated the usage pattern of “dǎ” by 30 Chinese-English bilingual preschool children in Singapore. Visual stimuli depicting “dǎ” actions were used to elicit descriptions from the participants. The results reveal that the meaning representations of “dǎ” in the semantic domains such as “social interaction” and “physical punishment” are most commonly used by the children while the meaning representations of “dǎ” in the semantic domains such as “fastening” and “possession” are the least used by the children. This paper will discuss the factors that affect the children’s use of the polysemous verb. MOE (Min. of Education, S’pore) Accepted version