Browsing by Issue Date, starting with "2003-11-16"

Now showing 1 - 20 of 72177

Fuzzy information retrieval based on multi-relationship fuzzy concept networks
(2003-11-16)
Chen, SM
;
Horng, YJ
;
Lee, CH
;
資訊工程學系
;
Department of Computer Science
In this paper, we present a new method for fuzzy information retrieval based on multi-relationship fuzzy concept networks. There are four kinds of fuzzy relationships in a multi-relationship fuzzy concept network, i.e., "fuzzy positive association" relationship, "fuzzy negative association" relationship, "fuzzy generalization" relationship and "fuzzy specialization" relationship. By performing fuzzy inferences based on the multi-relationship fuzzy concept network, the fuzzy information retrieval system can retrieve documents containing concepts that are not directly specified by the user but are somehow related to the user's query. In order to perform fuzzy inferences more efficiently, we use concept matrices to represent the degrees of fuzzy relationships between concepts in a multi-relationship fuzzy concept network. By calculating the transitive closures of concept matrices, the implicit degrees of fuzzy relationships between concepts are obtained. Multiple degrees of satisfaction that a document satisfies the user's query with respect to the fuzzy relationships between concepts are calculated. These satisfaction degrees are aggregated according to the user's specification to find the most relevant documents with respect to the user's query. The proposed fuzzy information retrieval method is more flexible and more intelligent than the one we presented in (IEEE Trans. Systems Man Cybernet.-Part B: Cybernet. 29(1) (1999) 126). (C) 2002 Elsevier B.V. All rights reserved.
Dielectric characteristics of low-permittivity silicate using electron beam direct patterning for intermetal dielectric applications
(2003-11-17)
Liu, PT
;
Chang, TC
;
Tsai, TM
;
Lin, ZW
;
Chen, CW
A direct patterning technology of low-permittivity silicate-based polymer is investigated with electron-beam lithography for multilevel interconnections. The smallest feature size of 60 nm for damascene lines can be directly patterned in the silicate film. In this direct patterning, dielectric regions exposed by electron beam are crosslinked and form desirable patterns, while the others are dissolvable in an aqueous solution containing 2.38% tetramethylammonium hydroxide. With an optimum condition of electron-beam lithography, the electron-beam-irradiated silicate exhibits superior dielectric properties than that of the furnace-cured silicate film, due to minimizing the break of Si-H bonds and moisture uptake. The explanation is in agreement with the analyses of Fourier transform infrared spectroscopy and thermal desorption spectroscopy. (C) 2003 American Institute of Physics.
Fast algorithm for evaluating the similarity of manufacturing processes within a dynamic production environment
(2003-11-20)
Huang, MG
;
Chang, PL
;
Chou, YC
;
經營管理研究所
;
Institute of Business and Management
To achieve efficient and economical production, planners and/or designers of production systems frequently must evaluate manufacturing process similarities for various products that they intend to produce. Such work is particularly important in product-mix production, which is widely used in various production systems, including product-oriented repetitive manufacturing, cellular manufacturing, and process-oriented manufacturing. A new potential application for evaluating manufacturing process similarity is necessary in dealing with capacity exchange between rush orders and prescheduled orders. Traditionally, most evaluating methods assumed an essentially static environment, and merely compared the manufacturing routing, processing time and demand. Meanwhile, other more complex evaluating methods were simply impractical. This study suggests a relatively simple but practical method for evaluating the manufacturing process similarity within a dynamic bucketed production environment. The proposed algorithm considers load projection and workstation utilization in terms of a current production schedule. Finally, a real numerical instance is cited to illustrate the effectiveness of the proposed approach.
Comment on "Triplet-to-singlet exciton formation in poly(p-phenylene-vinylene) light-emitting diodes" - Reply
(2003-11-21)
Lin, LC
;
Meng, HF
;
Shy, JT
;
Horng, SF
;
Yu, LS
Use WNx as diffusion barrier for copper airbridged low noise GaAsPHEMT
(2003-11-27)
Chang, HC
;
Chang, EY
;
Lien, YC
;
Chu, LH
;
Chang, SW
A low noise pseudomorphic high electron mobility transistor (PHEMT) with copper airbridges using sputtered WNx as the diffusion barrier has been developed. Both the material system and the copper airbridged PHEMT with WNx as the diffusion barrier did not decay even after thermal annealing at 250degreesC for 20 h. The results show that the copper airbridges with WNx diffusion barrier can be used as the interconnects for low noise GaAs PHEMTs.
Fault-tolerant cycle-embedding of crossed cubes
(2003-11-30)
Yang, MC
;
Li, TK
;
Tan, JJM
;
Hsu, LH
;
資訊工程學系
The crossed cube CQ(n) introduced by Efe has many properties similar to those of the popular hypercube. However, the diameter of CQ(n) is about one half of that of the hypercube. Failures of links and nodes in an interconnection network are inevitable. Hence, in this paper, we consider the hybrid fault-tolerant capability of the crossed cube. Letting f(e) and f(v) be the numbers of faulty edges and vertices in CQ(n), we show that a cycle of length 1, for any 4 less than or equal to l less than or equal to V(CQ(n)) - f(v) can be embedded into a wounded crossed cube as long as the total number of faults (f(v) +f(e)) is no more than n - 2, and we say that CQ(n) is (n - 2)-fault-tolerant pancyclic. This result is optimal in the sense that if there are n - 1 faults, there is no guarantee of having a cycle of a certain length in it. (C) 2003 Published by Elsevier B.V.
A model for the relative humidity effect on the readings of the PM10 beta-gauge monitor
(2003-12-01)
Chang, CT
;
Tsai, CJ
;
環境工程研究所
;
Institute of Environmental Engineering
In the previous paper (Atmos. Environ. 15 (1981) 1087), we found that the PM10 concentrations detected by the Wedding beta-gauge PM10 monitor and those measured by the manual hi-vol PM10 sampler were quite close when the ambient relative humidity (RH) was lower than the deliquescence RH (DRH) of aerosols. However, when the deliquescent point was exceeded, PM10 concentrations of the beta-gauge were found to be higher and differences increased with an increasing ambient RH. In addition, theoretical water mass calculated by a thermodynamic model (ISORROPIA model, (Aquat. Geochem. 4 (1998) 123)) was found to be much higher than the actual values. In this study, models were developed to determine water evaporation loss from collected particles on the filter tape of the beta-gauge during sampling and in the monitoring room. Simulated results show that all absorbed water will evaporate completely at RH lower than about 85%. However, absorbed water does not evaporate completely at RE higher than about 85%, and remaining water in particles accounts for higher beta-gauge readings than the hi-vol concentrations. The simulated daily beta-gauge PM10 concentrations are close to the actual beta-gauge readings obtained previously. (C) 2003 Elsevier Ltd. All rights reserved.
A method for measuring the complex refractive index and thickness of a thin metal film
(2003-12-01)
Chen, KH
;
Hsu, CC
;
Su, DC
;
光電工程學系
;
Department of Photonics
A circularly polarized heterodyne light beam is incident on a thin metal film, causing successive reflections and refractions to occur at the two sides of the thin film. The phase difference between p- and s-polarizations of the multiple-beam interference signal can be measured accurately with an analyzer and heterodyne interferometry. The phase difference depends on the azimuth angle of the analyzer, the complex refractive index and the thickness of the thin metal film. The measured values of the phase differences under three different azimuth angles of the analyzer can be substituted into the special equations derived from Fresnel's equations and multiple-beam interference. Hence, the complex refractive index and the thickness of the thin metal film can be estimated by using a personal computer with a numerical analysis technique. Because of its common-path optical configuration and its heterodyne interferometric phase measurement, this method has many merits, such as high stability against surrounding vibrations, high resolution and easy operation.
Effect of Shape and Size on Electron Transition Energies for Nanoscale InAs/GaAs Quantum Rings
(2003-12-01)
Li, Yiming
;
Lu, Hsiao-Mei
;
友訊交大聯合研發中心
;
D Link NCTU Joint Res Ctr
In this paper, we study the impact of the sizes and the shapes of nanoscale semiconductor quantum rings on the electron and hole energy states. A three-dimensional effective one band Schrodinger equation is solved numerically for semiconductor quantum rings with disk, cut-bottom-elliptical, and conical shapes. For small InAs/GaAs quantum rings we have found a sufficient difference in the ground state and excited state (l = -1) electron energies for rings with the same volume but different shapes. Volume dependence of the electron and hole energies can vary over a wide range and depends significantly on the ring shapes. It is found that a non-periodical oscillation of the energy band gap between the lowest electron and hole states as a function of external magnetic fields.
Metallurgical reaction of the Sn-3.5Ag solder and Sn-37Pb solder with Ni/Cu under-bump metallization in a flip-chip package
(2003-12-01)
Huang, CS
;
Duh, JG
;
Chen, YM
;
材料科學與工程學系
;
Department of Materials Science and Engineering
Several international legislations recently banned the use of Pb because of environmental concerns. The eutectic Sn-Ag solder is one of the promising candidates to replace the conventional Sn-Pb solder primarily because of its excellent mechanical properties. In this study, interfacial reaction of the eutectic Sn-Ag and Sn-Pb solders with Ni/Cu under-bump metallization (UBM) was investigated with a joint assembly of solder/Ni/Cu/Ti/Si(3)N(4)/Si multilayer structures. After reflows, only one (Ni,Cu)(3)Sn(4) intermetallic compound (IMC) with faceted and particlelike grain feature was found between the solder and Ni. The thickness and grain size of the IMC increased with reflow times. Another (Cu,Ni)(6)Sn(5) IMC with a rod-type grain formed on (Ni,Cu)(3)Sn(4) in the interface between the Sn-Pb solder and the Ni/Cu UBM after more than three reflow times. The thickness of the (Ni,Cu)(3)Sn(4) layer formed in the Sn-Pb system remained almost identical despite the numbers of reflow; however, the amounts of (Cu,Ni)(6)Sn(5) IMC increased with reflow times. Correlations between the IMC morphologies, Cu diffusion behavior, and IMC transformation in these two solder systems will be investigated with respect to the microstructural evolution between the solders and the Ni/Cu UBM. The morphologies and grain-size distributions of the (Ni,Cu)(3)Sn(4) IMC formed in the initial stage of reflow are crucial for the subsequent phase transformation of the other IMC.
Field test of a porous-metal denuder sampler
(2003-12-01)
Tsai, CJ
;
Huang, CH
;
Lin, YC
;
Shih, TS
;
Shih, BH
In this study, acidic/basic gas and particle concentrations determined with a porous-metal denuder and other samplers, that is, silica gel tube, impinger, honeycomb denuder system (HDS), Marple personal cascade impactor, and filter cassette, were compared in the ambient environment and different industrial sites in Taiwan. Results show that the concentrations determined by the denuder are very close to those of other samplers, with excellent correlation. The denuder also was tested successfully to determine H2SO4 size distribution in a lead-acid factory. One-way ANOVA shows no significant differences (P > 0.05) between the porous-metal denuder (PDS) and other samplers. The denuder is compact in size and suitable for sampling in the workplace and ambient environment.
Fully silicided NiSi and germanided NiGe dual gates on SiO2 n- and p-MOSFETs
(2003-12-01)
Yu, DS
;
Wu, CH
;
Huang, CH
;
Chin, A
;
Chen, WJ
We have fabricated the fully silicided NiSi and germanided NiGe dual gates n- and p-MOSFETs on 1.9 nm thick SiO2 gate dielectric. The extracted work functions of fully NiSi and NiGe gates from thickness-dependent flat band voltage were 4.55 and 5.2 eV respectively, which may provide possible wide work function tuning using NiSi1-xGex. In additional to the lower gate current than Al gate n- and p-MOSFETs, the fully silicided NiSi and germanided NiGe gates MOSFETs show electron and hole mobilities close to universal mobility values with special advantage of process compatible to current VLSI fabrication line.
Effects of underlayer dielectric on the thermal characteristics and electromigration resistance of copper interconnect
(2003-12-01)
Chin, YL
;
Chiou, BS
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
The effects of underlayer dielectric on the thermal characteristics and electromigration resistance of copper interconnects are investigated. It is shown that the thermal impedance of metal lines on polyimide is about three times that of metal lines on plasma-enhanced chemical vapor deposition of tetraethylorthosilicate (PETEOS) oxide. The open-circuit failure occurs after 75 s stressing at a current density of 1.4 x 10(7) A/cm(2) for polyimide underlayer specimens due to the high joule heating effect which causes the extreme temperature rise at, interconnects. Hence, decomposition of polyimide occurred which further accelerated the breakdown of interconnects. The thermal impedance is one of the major factors causing the electrical resistance to change during the electromigration (EM) test. The activation energies for electromigration of Cu are 0.87 eV for SiO2/Cu/PI2610 and 0.65 eV for SiO2/Cu/SiO2. After 500 thermal cycles, both the mean time-to-failure (MTF) and activation energy E-a decrease for EM, and lognormal standard deviation increases. At 225degreesC, the MTF of SiO2/Cu/PI2610 is about two times longer than that of SiO2/Cu/SiO2.
STRUCTURES AND PROPERTIES OF THE CRYSTALLINE Si-C-N USING ADDITIONAL Si-SOURCE AND Co-CATALYST
(2003-12-01)
Chang, Hui Lin
;
Fang, Jau Shiung
;
Kuo, Cheng Tzu
;
材料科學與工程學系
;
Department of Materials Science and Engineering
Si-C-N films were synthesized on Si substrates by microwave plasma chemical vapor deposition (MPCVD) with a mixture of CH4 and N-2 as gaseous sources, and Co-coated Si columns to act as a catalyst and an additional Si source. Two conditions, conditions 1 and 2, were implemented by varying the time at which the solid sources were introduced. Under condition 1, the solid sources were applied \'before\' film deposition. Under condition 2, the solid sources were applied both \'before and during\' film depositions. Analytical results indicate, although that both conditions yield crystalline Si-C-N films, the time at which the solid sources are applied is critically affects the films\' structures and properties. Synthesizing Si-C-N films under condition 1 yields crystals with more re-nucleation, a structure closer to a pseudo T-Si3N4 structure, a higher C content, detectable Si(2p)-C bonding, lower nano-hardness and better field emission properties. Synthesis of film under condition 2 yields, crystals with more facets, a structure closer to pseudo alpha-Si3N4 and an additional thin layer under the crystalline layer. SEM, TEM, XPS, XRD, nanoindentor and field emission were used to characterize the film\' structures and properties.
Derivation of variable soft sphere model parameters in direct-simulation Monte Carlo method using quantum chemistry computation
(2003-12-01)
Wu, JS
;
Hsu, YL
;
機械工程學系
;
Department of Mechanical Engineering
A practical procedure for deriving parameters of the variable soft sphere (VSS) molecular model for the Lennard-Jones potential of gas species in the direct-simulation Monte Carlo (DSMC) method is introduced. Intermolecular potential energy is fitted through the Lennard-Jones (L-J) potential, which is obtained by quantum chemistry computation. Computed L-J potentials are then used to compute viscosity and diffusivity coefficients for like and unlike molecules, respectively, based on gas kinetics theory. Related parameters of the VSS molecular model are derived by fitting these computed coefficients to those derived from the VSS model. Test results for O-2-H-2 Mixtures using this procedure are in good agreement with existing data. VSS parameters for silicon-based species frequently encountered in the chemical vapor deposition process are derived as an example to illustrate the importance of this procedure in extending the applicability of the DSMC computation to gas species other than air-related species.
Synthesis and characterization of fluorinated polybenzoxazine material with low dielectric constant
(2003-12-01)
Su, YC
;
Chang, FC
;
應用化學系
;
Department of Applied Chemistry
It is well-known that low dielectric materials are used as insulating materials, and the incorporation of fluorinated substitutes into polymer is able to decrease its dielectric constant because of small dipole and the low polarizability of the C-F bond. In this study, a novel structure of the fluorinated benzoxazine (F-I benzoxazine) has been synthesized by incorporating the trifluoromethyl groups into the monomer, and its structure has also been characterized by H-1 NMR, F-19 NMR and FT-IR. Further, we have prepared the fluorinated copolybenzoxazine (B-a/F- = 1/1) with substantially reduced dielectric constant at K = 2.36. In addition, this fluorinated copolybenzoxazine possesses high glass transition temperature and high thermal stability, which is suitable for high temperature operation for certain special processes of interlayer dielectrics. (C) 2003 Published by Elsevier Ltd.
Two-stage L-band EDFA applying C/L-Band wavelength-division multiplexer with the counterpropagating partial gain-clamping
(2003-12-01)
Lu, YH
;
Chi, SE
;
光電工程學系
;
Department of Photonics
A simple ring laser technique of two-stage gain-clamped long-wavelength band erbium-doped fiber amplifier is investigated. The ring laser light is excluded at the output end and noise-figure degradation can. be reduced due to the counterpropagating partial gain-clamping in the second stage. The laser mechanism is achieved by the C/L-band wavelength-division multiplexer instead of a tunable bandpass filter and an optical circulator at the output end, which,were often used in the previous counterpropagating ring laser architectures, and thus, this design is cost-effective. The gain-clamping of long-wavelength band (1569-1606.2 nm) in our configuration is demonstrated experimentally for the saturating tone (1587.6 nm) power up to -2 dBm.
Internet addiction of adolescents in Taiwan: An interview study
(2003-12-01)
Tsai, CC
;
Lin, SSJ
;
師資培育中心
;
Center of Teacher Education
This study presented a follow-up investigation of a series of studies about Taiwanese adolescents' Internet addiction. Through analyzing a series of questionnaires, 10 Taiwanese adolescents with Internet addiction were selected for in-depth interviews. Adolescents' interview data revealed the following: First, almost all of the interviewed adolescents exhibited most of the symptoms of Internet addiction as identified by previous studies, including compulsive use and withdrawal, tolerance, and related problems of school, health, family, finance, and time management. Moreover, almost all of the interviewed adolescents stated that messages on the Internet became the primary source of their information and knowledge. Many of them were addicted to the messages and activity on the Internet, but not by the Internet as a medium per se. Also, the world of the Internet seemed to become a place to relieve adolescents' depression. Most of the interviewed adolescents self-declared that they were Internet addicts, but none of them could clearly state what to do with such an addiction. Psychologists and educators should pay more attention to this and then propose possible solutions for adolescents.
Implications of the Dirichlet assumption for discretization of continuous variables in naive Bayesian classifiers
(2003-12-01)
Hsu, CN
;
Huang, HJ
;
Wong, TT
;
資訊工程學系
;
Department of Computer Science
In a naive Bayesian classifier, discrete variables as well as discretized continuous variables are assumed to have Dirichlet priors. This paper describes the implications and applications of this model selection choice. We start by reviewing key properties of Dirichlet distributions. Among these properties, the most important one is "perfect aggregation," which allows us to explain why discretization works for a naive Bayesian classifier. Since perfect aggregation holds for Dirichlets, we can explain that in general, discretization can outperform parameter estimation assuming a normal distribution. In addition, we can explain why a wide variety of well-known discretization methods, such as entropy-based, ten-bin, and bin-log l, can perform well with insignificant difference. We designed experiments to verify our explanation using synthesized and real data sets and showed that in addition to well-known methods, a wide variety of discretization methods all perform similarly. Our analysis leads to a lazy discretization method, which discretizes continuous variables according to test data. The Dirichlet assumption implies that lazy methods can perform as well as eager discretization methods. We empirically confirmed this implication and extended the lazy method to classify set-valued and multi-interval data with a naive Bayesian classifier.
Hierarchical learning architecture with automatic feature selection for multiclass protein fold classification
(2003-12-01)
Huang, CD
;
Lin, CT
;
Pal, NR
;
電控工程研究所
;
Institute of Electrical and Control Engineering
The structure classification of proteins plays a very important role in bioinformatics, since the relationships and characteristics among those known proteins can be exploited to predict the structure of new proteins. The success of a classification system depends heavily on two things: the tools being used and the features considered. For the bioinformatics applications, the role of appropriate features has not been paid adequate importance. In this investigation we use three novel ideas for multiclass protein fold classification. First, we use the gating neural network, where each input node is associated with a gate. This network can select important features in an online manner when the learning goes on. At the beginning of the training, all gates are almost closed, i.e., no feature is allowed to enter the network. Through the training, gates corresponding to good features are completely opened while gates corresponding to bad features are closed more tightly, and some gates may be partially open. The second novel idea is to use a hierarchical learning architecture (HLA). The classifier in the first level of HLA classifies the protein features into four major classes: all alpha, all beta, alpha + beta, and alpha/beta. And in the next level we have another set of classifiers, which further classifies the protein features into 27 folds. The third novel idea is to induce the indirect coding features from the amino-acid composition sequence of proteins based on the N-gram concept. This provides us with more representative and discriminative new local features of protein sequences for multiclass protein fold classification. The proposed HLA with new indirect coding features increases the protein fold classification accuracy by about 12%. Moreover, the gating neural network is found to reduce the number of features drastically. Using only half of the original features selected by the gating neural network can reach comparable test accuracy as that using all the original features. The gating mechanism also helps us to get a better insight into the folding process of proteins. For example, tracking the evolution of different gates we can find which characteristics (features) of the data are more important for the folding process. And, of course, it also reduces the computation time.