University of Danang
University of Danang
My L.H.T.,University of Danang |
Huy K.P.,Danang University of Technology
International Conference on Advanced Technologies for Communications | Year: 2016
All operations of the information processing on computer relate to text editor. The misspellings or typing errors often make in text editor. In MS Office applications have English spellchecking function. The current Vietnamese spellchecking methods are mostly based on Vietnamese vocabulary corpus because most words are presented in these corpus. Currently, the Ede vocabulary corpus is also being gradually improved. Therefore, Ede spellchecking basing on vocabulary corpus has a lot of difficulties. From that, the spellchecking Ede syllable basing on the Ede syllable models has been proposed. The objective of this study is building the Ede syllable spellchecking algorithm basing on the Ede syllable models. This algorithm is applied to separate the words which are not in the Ede vocabulary into two structure groups: correct and incorrect syllables. The correct syllables are saved for the input data of the Ede-Vietnamese vocabulary corpus developing application. This research has contributed to support spellchecking Ede language in text editor and to develop Ede and Ede-Vietnamese vocabulary corpus. © 2016 IEEE.
Nguyen-Duy-nhat V.,University of Danang |
Bui-Thi-minh T.,University of Danang |
Tang-Tan C.,University of Danang |
Baa V.N.Q.,Posts and Telecommunications Institute of Technology |
Nguyen-Le H.,Sudan University of Science and Technology
International Conference on Advanced Technologies for Communications | Year: 2016
This paper is concerned with the joint estimation of phase noise and time-varying channel responses in full-duplex multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems. The time-variation of phase noise and multipath MIMO channel responses gives rise to the identifiability problem of estimating a large number of unknowns which is much greater than the number of received signal samples. To overcome the resulting identifiability estimation problem, this paper employs basis expansion models (BEM) for reducing the number of time-varying channel impairment parameters to be estimated. Using the BEM-based approximation of the channel impairment responses, this paper formulates a pilot-Aided algorithm for maximum-A-posteriori (MAP)-based estimation of the BEM coefficients. This paper provides empirical mean squared error (MSE) results of the joint channel impairment parameter estimation and related analytical values of Bayesian Cramer Rao lower bounds (BCRLB). The numerical results illustrate that the use of BEM facilitates reliable estimation of the time-varying channel impairment responses whereas the corresponding MSE and bit-error-rate (BER) values are comparable to related BCRLBs and ideal BER ones, respectively. © 2016 IEEE.
Omar H.,National Chiao Tung University |
Omar H.,Industrial Technology Research Institute of Taiwan |
Hoang V.H.,University of Danang |
Liu D.-R.,National Chiao Tung University
Computational Intelligence and Neuroscience | Year: 2016
Enhancing sales and operations planning through forecasting analysis and business intelligence is demanded in many industries and enterprises. Publishing industries usually pick attractive titles and headlines for their stories to increase sales, since popular article titles and headlines can attract readers to buy magazines. In this paper, information retrieval techniques are adopted to extract words from article titles. The popularity measures of article titles are then analyzed by using the search indexes obtained from Google search engine. Backpropagation Neural Networks (BPNNs) have successfully been used to develop prediction models for sales forecasting. In this study, we propose a novel hybrid neural network model for sales forecasting based on the prediction result of time series forecasting and the popularity of article titles. The proposed model uses the historical sales data, popularity of article titles, and the prediction result of a time series, Autoregressive Integrated Moving Average (ARIMA) forecasting method to learn a BPNN-based forecasting model. Our proposed forecasting model is experimentally evaluated by comparing with conventional sales prediction techniques. The experimental result shows that our proposed forecasting method outperforms conventional techniques which do not consider the popularity of title words. © 2016 Hani Omar et al.
Giang L.T.,Danang People Committee |
Hung V.T.,University of Danang |
Phap H.C.,University of Danang
ACM International Conference Proceeding Series | Year: 2013
Using bilingual dictionaries is a common way for query translation in Cross Language Information Retrieval. In this article, we focus on Vietnamese-English Bilingual Information Retrieval and present algorithms for query segmentation, word disambiguation and re-ranking to improve the dictionary-based query translation approach. An evaluation environment is implemented to verify and compare the application of proposed algorithms with the baseline method using manual translation. Copyright © 2013 ACM.
Nguyen L.-N.,Academia Sinica, Taiwan |
Nguyen L.-N.,University of Danang |
Lan Y.-W.,Academia Sinica, Taiwan |
Chen J.-H.,Chung Yuan Christian University |
And 6 more authors.
Nano Letters | Year: 2014
Two-dimensional crystals can be assembled into three-dimensional stacks with atomic layer precision, which have already shown plenty of fascinating physical phenomena and been used for prototype vertical-field-effect- transistors.1,2 In this work, interlayer electron tunneling in stacked high-quality crystalline MoS2 films were investigated. A trilayered MoS2 film was sandwiched between top and bottom electrodes with an adjacent bottom gate, and the discrete energy levels in each layer could be tuned by bias and gate voltages. When the discrete energy levels aligned, a resonant tunneling peak appeared in the current-voltage characteristics. The peak position shifts linearly with perpendicular magnetic field, indicating formation of Landau levels. From this linear dependence, the effective mass and Fermi velocity are determined and are confirmed by electronic structure calculations. These fundamental parameters are useful for exploitation of its unique properties. © 2014 American Chemical Society.
Belec L.,University of Toulon |
Nguyen T.H.,University of Danang |
Nguyen D.L.,Sudan University of Science and Technology |
Chailan J.F.,University of Toulon
Composites Part A: Applied Science and Manufacturing | Year: 2015
The effects of humid tropical conditions and artificial ageing tests on a UD glass-fibre/epoxy composite are compared at different scales. A modified network of 500 nm width around each fibre is identified by AFM force measurements. The evolution of this interphase is correlated to the composite embrittlement observed in transverse bending tests during ageing. DSC and DMA show up the matrix plasticization under tropical conditions and hygrothermal ageing. Chain scissions are also involved under artificial and natural UV exposure on surface layers and photoproducts are identified by FTIR analysis. The effects of artificial and natural hygrothermal conditions are emphasised inside interphases. The interfacial areas situated close to the composite surface are highly degraded by UV radiation. Cracks can then be initiated in these areas during bending tests. Finally, AFM force measurements highlight synergistic effects during natural weathering due to the combined effects of UV radiation and hygrothermal conditions. © 2014 Elsevier Ltd. All rights reserved.
Ngo T.C.,Duy Tan University |
Dao D.Q.,Duy Tan University |
Thong N.M.,University of Danang |
Nam P.C.,The University of Da nang
RSC Advances | Year: 2016
The antioxidant properties of 21 non-phenolic terpenoids contained in essential oil extracted from the buds of Cleistocalyx operculatus have been investigated using density functional theory (DFT)-based computational methods. The C-H bond dissociation enthalpy (BDE), proton dissociation enthalpy (PDE), proton affinity (PA), electron transfer enthalpy (ETE) and ionization energy (IE) were calculated in the gas phase and in two different solvents (water and ethanol) at the ROB3LYP/6-311++G(2df,2p)//B3LYP/6-311G(d,p) level of theory for the former and with the PM6 method for the latter. Quantum chemical descriptors like chemical potential (μ), chemical hardness (η), and global electrophilicity (ω) were calculated in order to evaluate the reactivity and stability of all studied compounds. The interaction of the HOO. radical with α-terpinene, an example molecule, was also studied in detail by establishing a potential energy surface (PES). As a result, a kinetic concurrence between the H-abstraction reaction at the weakest C-H bond and addition reactions at the C=C double bonds was clarified. On the basis of this mechanism, the antioxidant capacity may happen via the termination-enhancing process. Among the studied compounds, α-terpinene, γ-terpinene, cembrene and abieta-7,13-diene represent potential antioxidants. © The Royal Society of Chemistry 2016.
Nguyen D.S.,University of Danang
ASME International Mechanical Engineering Congress and Exposition, Proceedings (IMECE) | Year: 2013
Due to development of science and information technology, computer has become an effective and useful tool supporting for product design activities. The numerical model of product is quickly created in Computer-Aided-Design (CAD) environment by product designers. However, this model is a nominal model representing ideal geometric information of product. It is not able to cope with various kinds of disturbances during the whole product life-cycle, especially in manufacturing stage where geometric deviations of product are generated and accumulated by material property defects, manufacturing process errors, tooling deformation, etc. This is a limitation of the current product modeling technology. In addition, most of the simulations of product performance such as kinematics, dynamics, failure, etc., are carried out by using this nominal model. It can make the quality of the product designed not to fully meet the requirements of customers and users. Thus, the paper proposes a method that allows integrating geometric deviations of product in CAD environment. The aim is to deal with the above issues and to take into account various disturbances in product performance prediction. Copyright © 2013 by ASME.
Truong C.D.,Hanoi University of Science and Technology |
Quyen N.X.,Hanoi University of Science and Technology |
Hung P.V.,Vietnam Maritime University |
Vu T.H.,University of Danang
2016 IEEE 6th International Conference on Communications and Electronics, IEEE ICCE 2016 | Year: 2016
In this paper, we present a novel mode synthesizer based on a silicon-on-insulator (SOI) platform with the advantages of broad bandwidth, high mode conversion efficiency and large fabrication tolerance. By using a 3×1 multimode interference (MMI) coupler, a sinusoidal symmetric Y junction coupler and two linear butterfly-shaped phase shifters, mode synthesis from the fundamental mode TE0 and the first-order mode TE1 to the second-order mode TE2 is achieved by means of numerical simulation. Phase variation in this structure is discussed and performance analysis of the mode synthesizer is performed by means of the numerical method. Simulation results obtained by using three dimension beam propagation method (3D-BPM) incorporated with effective index method (EIM) show high performance of the device with mode conversion efficiency up to more larger than 88% (corresponding to insertion loss less than 0.52 dB) in the whole C-band. The width and length tolerances of the multimode interference coupler are quite large, which can be corresponding up to 50 nm and 900 nm, respectively. In addition, the length deviation of the phase shifters can be up to 500 nm. Large tolerances allow to fabricate the device by using current CMOS technology. © 2016 IEEE.
PubMed | University of Danang
Type: Journal Article | Journal: The journal of physical chemistry. B | Year: 2016
We have investigated the interactions between polygalacturonate (polyGal) and four divalent cations (M(2+) = Ba(2+), Ca(2+), Mg(2+), Zn(2+)) that differ in size and affinity for water. Our results evidence that M(2+)-polyGal interactions are intimately linked to the affinity of M(2+) for water. Mg(2+) interacts so strongly with water that it remains weakly bound to polyGal (polycondensation) by sharing water molecules from its first coordination shell with the carboxylate groups of polyGal. In contrast, the other cations form transient ionic pairs with polyGal by releasing preferentially one water molecule (for Zn(2+)) or two (for Ca(2+) and Ba(2+)), which corresponds to monodentate and bidentate binding modes with carboxylates, respectively. The mechanism for the binding of these three divalent cations to polyGal can be described by two steps: (i) monocomplexation and formation of point-like cross-links between polyGal chains (at low M(2+)/Gal molar ratios, R) and (ii) dimerization (at higher R). The threshold molar ratio, R*, between these two steps depends on the nature of divalent cations and is lower for calcium ions (R* < 0.1) than for zinc and barium ions (R* > 0.3). This difference may be explained by the intermediate affinity of Ca(2+) for water with respect to those of Zn(2+) and Ba(2+), which may induce the formation of cross-links of intermediate flexibility. By comparison, the lower and higher flexibilities of the cross-links formed by Zn(2+) and Ba(2+), respectively, may shift the formation of dimers to higher molar ratios (R*).