34254 research outputs found
Sort by
Analyzing Online Reviews to Explore Dining Experience Differences Before and After Recognition: A Case Study of Bib Gourmand Restaurants in Taipei
No full textWith the introduction of the Michelin Guide in Taipei in 2018, it has gained a certain level of influence and impact on the dining market as well as consumers in recent years. When choosing restaurants, consumers can refer to the recommendations of the Michelin Guide. Along with the rise of online reviews, it has also influenced consumer purchasing decisions. Consumers rely on the content of online reviews as a reference for their choices.
This study is based on the reviews of restaurants in Taipei that were newly selected in the Michelin Guide Bib Gourmand Award from 2020 to 2022. It collected over 7,000 online reviews from the year before their selection and the year after their selection on Google Maps. Through text analysis techniques, the collected reviews were subjected to thematic analysis, sentiment analysis, and word frequency analysis. The aim was to observe the differences in consumer perception of the Michelin-selected restaurants before and after their selection.
Based on dining experiences, this study categorizes the review content into four dimensions: "food," "service," "price," and "ambiance," using topic modeling to analyze online reviews. The research intends to identify the significant differences among these four dimensions of the dining experience. Consumers' reviews show a higher sensitivity towards the "food" dimension compared to the other three dimensions.
Our study shows that most reviews primarily focus on "food" and "price," suggesting that consumers have certain expectations and standards for food quality and pricing. Through the analysis of online review content, it is observed that consumers' negative emotions mainly stem from the "food" and "service" dimensions.
From a managerial perspective, this study suggests that catering establishments should focus on enhancing food quality to meet consumers' expectations for their dining experience. Additionally, improving service attitude and efficiency while minimizing negative evaluations are recommended. The findings of this study aim to assist catering establishments to meet consumer needs. Through detailed analysis of sentiment scores, targeted improvement measures can be implemented for different dimensions, thereby providing better products and services. Ultimately, this can enhance the dining experience for consumers and promote the overall development of the catering market
Numerical Modeling on Wave Propagation in the Nearshore Zone Using Smoothed Particle Hydrodynamics Method
No full textThe numerical simulation software used in this study is DualSPHysics. This software employs the Smoothed Particle Hydrodynamics (SPH) method, a Lagrangian meshless approach. It has significant advantages in solving complex physical problems such as free surface flow phenomena, wave-structure interactions, and multiphysics coupling.Since the development history of this software is relatively short, the main focus of this research is to test the accuracy of DualSPHysics in simulating wave propagation on different types of seabed, as well as phenomena like shoaling and wave breaking. Additionally, the software is applied to simulate wave propagation on sediment sloping seabed and fixed sloping seabed, as well as the generation of storm surges caused by long waves on the northeastern coast of Taiwan. The results of this study demonstrate that DualSPHysics can effectively represent the physical phenomena of wave generation in a still water tank, wave propagation behavior on slopes, and nonlinear changes when waves pass through underwater structures. Regarding the variations of wave propagation on fixed and varying seabeds, it is observed that the influence of varying seabeds mainly occurs before and after wave breaking, and sediment deposition on the seabed leads to greater dissipation of wave energy during wave breaking.Lastly, in terms of wave generation along the northeastern coastal, it is found that the likelihood and intensity of wave generation are closely related to the slope of nearshore topography. The steeper the slope, the more likely it is to generate storm surges. It canalso be observed that waves surging onto the platform are accompanied by higher horizontal flow speeds with substantial energy, making it easy for more dangerous storm surges to occur
Optimal Phase-Tracking Self-Injection-Locked Radar for Through-Wall Life Detection
No full textIn recent years, microwave radar has been widely used in the field of vital sign detection, with continuous wave radar being the most commonly used type. To enhance sensing capabilities, self-injection-locked radar has been developed, which offers superior sensitivity compared to traditional continuous wave radar, enabling long-range detection of subtle chest movements caused by cardiopulmonary activities. However, self-injection-locked radar is still affected by the detection null point, and in the presence of clutter, frequency offset can impact the sensing range and sensitivity of the radar system. To address these issues, this paper proposes the use of an optimal phase tracking self-injection-locked radar to overcome the problems of detection null point and frequency offset.
This paper also analyzes and measures the performance of the optimal phase tracking self-injection-locked radar, utilizing both self-injection-locked radar and frequency-locked loop techniques, to improve the sensing efficiency and accuracy of the radar. The combination involves adding a frequency-locked loop at the demodulation end of the self-injection-locked radar, ensuring the phase of the self-injection-locked loop remains at 180 degrees, thus maintaining optimal self-injection locking phase. Additionally, this study employ a phase shifter to eliminate the influence of clutter, thereby enhancing the sensing capability of the radar system in different environments. After formula derivation and analysis, the proposed optimal phase tracking self-injection-locked radar system is validated through experiments. In the experiments, the optimal phase tracking self-injection-locked radar is used to monitor the vital signs of a subject located 3 meters behind a concrete wall. The results demonstrate that the radar system can successfully and accurately detect the vital sign of the subject beyond the concrete wall
Synthetic Aperture Self-Injection-Locked Radar with Clutter Suppression and 3-D Positioning Capabilities for Biomedical Application
No full textThis thesis presents a Synthetic Aperture Radar (SAR) based on self-injection-locking (SIL) technology. The SIL radar was mounted on a constant-speed platform to perform SAR measurements. Then, the 2D imaging and target location are obtained using range Doppler algorithm (RDA). Moreover, the Doppler information of the targets can be extracted from the slow-time phase measured by the high-sensitivity SIL radar. This study preserves the images of moving targets but removes those of static objects in the SAR measurements by using the targets\ue2 Doppler information. This study not only conducts experiments to provide the positions and images of the single or multiple targets under different indoor conditions, but also demonstrates the through-wall imaging by placing a wood wall between the targets and radar.
This system also operates as an Interferometric synthetic aperture radar (InSAR). The height between the targets and radar can be calculated in terms of the slow-time phase difference between two radar trajectories, and the actual height of the targets can be estimated from adding the height of the radar. This thesis finally demonstrates the capability of 3D positioning and using the information provided from InSAR and 2D SAR measurements
Development of a Rapid Response Micro Gas Sensor with a Suspended Micro Heater for Methyl Mercaptan Gas Detection
No full textMethyl mercaptan is a colorless, flammable gas with a characteristic rotten cabbage odor. It is commonly found in chemical, petrochemical, and agricultural industries. The olfactory threshold for methyl mercaptan is 41 ppb, and according to the Occupational Safety and Health Administration (OSHA) in the United States, the permissible exposure limit (PEL-Ceiling) for methyl mercaptan is 10 ppm. Prolonged exposure to an environment containing methyl mercaptan can lead to olfactory fatigue, rendering individuals unable to perceive its presence, thereby increasing the risk of accidents. To mitigate such unfortunate incidents, this research aims to assist industries in developing micro gas sensors for methyl mercaptan detection with rapid response and recovery characteristics.
In this study, microelectromechanical systems (MEMS) technology was employed to develop a micro gas sensor for methyl mercaptan detection. A micro heater was integrated into the sensing chip to provide the optimal operating temperature for the sensor, as well as to significantly reduce the device size for future microsystem applications. Furthermore, to effectively minimize the power consumption of the heater, silicon substrates were selectively etched using KOH to create a backside cavity, and the front-side film was etched using inductively coupled plasma (ICP) to fabricate a suspended structure for the micro gas sensor and heater, thereby reducing silicon's thermal conductivity. Additionally, different geometric designs of the heating electrodes were implemented and optimized to improve heating uniformity and further reduce power consumption. Finally, a tin oxide (SnO2) thin film deposited by sputtering was utilized as the sensing layer for methyl mercaptan to enhance the sensor response, shorten response and recovery times.
The developed methyl mercaptan sensing chip in this study has dimensions of 5 mm \uc3 5 mm and incorporates a suspended micro heater. The micro heater can be heated to 300\uc2\ub0C with a voltage of only 5.57 V and a power of 172 mW. To determine the optimal operating temperature for the methyl mercaptan sensing chip, the device characteristics were measured at five different temperatures ranging from 150\uc2\ub0C to 250\uc2\ub0C. The results revealed that the best operating temperature was 200\uc2\ub0C, requiring only 4.47 V and 113 mW of power. The concentration range of methyl mercaptan measured in this study was 40 ppb to 25 ppm. Comparing the characteristics of the sensor at six different concentrations, it was observed that the sensor response increased with increasing gas concentration. Even at the lowest test concentration of 40 ppb, a response of 1.21 was achieved, with a rapid response time of 15 seconds and a recovery time of 6 seconds. At the maximum permissible concentration of 10 ppm for human exposure, a response of 3.73 was obtained, with a response time of 3 seconds and a recovery time of 15 seconds.In summary, the micro methyl mercaptan sensing chip developed in this study offers advantages such as rapid response and recovery times, low detection limit, low power consumption, and miniaturized size
Hospice nurses\ue2 perspectives and challenges on holistic care in multiple institutions.
No full textBackground: Holistic care is a comprehensive care model, including the body, mind, spirit and society. Palliative care is part of holistic care, improving patients' quality of life, and reducing overuse medical care. Nurses are the primary group providing holistic care. With the increasing burden of aging and non-communicable diseases, the demand for palliative care has grown significantly.
Purposes: Taiwan's growing emphasis on palliative care and many foreign studies have shown that nurses\ue2 understanding and implementation of holistic care are relatively weak. Therefore, this study aims to understand hospice nurses' perceptions and behaviors of holistic care and the challenges or difficulties encountered in implementation to identify the difficulties and help make the work smoother.
Methods: The subjects of this study were hospice nurses. In order to deeply understand their views and behaviors, the study was divided into two phases of questionnaire design and qualitative interview according to different purposes. The samples were from a southern medical center and its cooperative district hospitals, regional hospitals, and medical centers in a cross-hospital and multi-institutional manner. The knowledge, attitude, and behavior theory were used for analysis.
Results: The questionnaire results indicate that knowledge, attitude, and behavior mutually influence each other, with attitude acting as a complete mediator. Additionally, receiving educational training has a significant impact on both attitude and behavior. In the interviews, the respondents' perspectives and behaviors were quite similar. Furthermore, the foremost challenge identified was related to the patient\ue2s family members.
Conclusions: Through research to understand hospice nurses' perceptions and behaviors on holistic care, find out the difficulties and challenges encountered in implementation, and establish appropriate education training and management systems, which will promote nurses to provide holistic care and improve the quality of care
Robust Transmission for Energy-Efficient Sub-Connected Active RIS-Assisted Wireless Networks: DRL versus Traditional Optimization
No full textThis study explores the effectiveness of a sub-connected active reconfigurable intelligent surface (RIS) in communication scenarios with imperfect channel state information (CSI). The main objective is to ensure reliable transmission by formulating an optimization problem that maximizes energy efficiency (EE). This involves jointly optimizing the transmit precoder at the base station (BS) and the beamforming matrix at the RIS, considering the norm-bounded CSI error model. As the problem is non-convex, we employ deep reinforcement learning (DRL) techniques, specifically deep deterministic policy gradient (DDPG), proximal policy optimization (PPO), and modified PPO approaches to get the optimal transmit precoder and
beamforming matrix. The proposed algorithms\ue2 convergence and robustness are demonstrated through extensive simulations. Furthermore, we analyze the consequence of key system parameters, such as
the total number of elements, the required number of amplifiers, and the maximum available transmit power at the BS, on the system\ue2s performance. In conclusion, we compare the proposed algorithms with traditional analytical optimization methods to validate their efficacy
A New Line Tunneling SiGe/Si iTFET with Control Gates for Leakage Suppression and Subthreshold Swing Improvement
No full textThis thesis proposes a new line tunneling dominating metal-semiconductor contact-induced SiGe tunnel field-effect transistor with controlled gates(CG-Line iTFET), as well as a line tunneling dominating metal-semiconductor contact-induced SiGe/Si tunnel field-effect transistor with controlled gates(CG-Line SiGe/Si iTFET). The main difference between the two design architectures lies in the body region: the former consists of a single material, Si0.7Ge0.3, while the latter is divided into a source region of Si0.7Ge0.3 and a drain region of Si. In the CG-Line iTFET design, both drain ends have symmetrical control gates. When a sufficiently negative fixed bias is applied to the control gates, the overlap of the energy bands at the drain region in the OFF-state can be effectively suppressed, reducing carrier tunneling probability and significantly decreasing leakage current. Additionally, due to the large overlap area between the source and gate, the gate can effectively control the tunneling junction, resulting in a more uniform distribution influenced by the gate's electric field. Compared to conventional tunnel field-effect transistors dominated by point tunneling, this design enhances the ON-state current and subthreshold swing characteristics. In the CG-Line iTFET, we use N-type bulk SiGe as the body region of the device. By using metal-semiconductor contacts with different work functions, the junctions can form either ohmic contacts or Schottky contacts. Here, we utilize the characteristics of Schottky contacts, which induce an opposite type of doping concentration on the surface of the N-type bulk material, forming a PN junction. This approach avoids issues related to doping profiles and random doping fluctuations and reduces the thermal budget by minimizing ion implantation and subsequent annealing processes. By optimizing the geometry and experimental parameters of the CG-Line iTFET, various electrical characteristics of the device can be improved. For the proposed CG-Line iTFET, at a drain bias voltage (VDS) of 0.3 V, we achieve a minimum subthreshold swing (SSmin) of 4.77 mV/dec, and an average subthreshold swing (SSavg) of 24.48 mV/dec. The ON-state current (ION) is 2.36 \uc3 10-6 A/\uce\ubcm, and the ION / IOFF ratio is 5.47 \uc3 107. To further reduce the OFF-state current, we introduce the second architecture, CG-Line SiGe/Si iTFET. In this design, the body region is divided into a source region and a drain region, with different bandgap materials. The source region utilizes Si0.7Ge0.3 with a smaller bandgap, increasing the probability of band-to-band tunneling in the ON-state and improving the ON-state current. The drain region employs Si with a larger bandgap, reducing the probability of band-to-band tunneling in the OFF-state and thus lowering the OFF-state leakage. By appropriately adjusting the doping concentration in the source region, a significant reduction in OFF-state current can be achieved due to the reduction in leakage current paths. The CG-Line SiGe/Si iTFET device achieves a minimum subthreshold swing (SSmin) of 3.58 mV/dec and an average subthreshold swing (SSavg) of 19.12 mV/dec at a bias voltage (VDS) of 0.3 V. The ON-state current (ION) is 2.26 \uc3 10-6 A/\uce\ubcm, and the ION / IOFF ratio is 1.22 \uc3 109. In future integrated circuit devices, to improve performance and extend the standby time of electronic products, the primary requirements are low power consumption and fast switching characteristics. As the supply voltage gradually decreases, a steep subthreshold swing becomes crucial. Even with a slight change in gate bias voltage, the current can be greatly enhanced, providing sufficient ON-state current to drive the circuit. Therefore, the focus of this master's thesis is primarily on enhancing the device's ON-state current and achieving a steeper subthreshold swing, with the aim of applying it to future Internet of Things (IoT), Artificial Intelligence (AI), and various low-power electronic products
Range Comparable Attribute-Based Encryption Scheme with Partially Hidden Access Policies
No full textAs the amount of data uploaded to the cloud grows rapidly, the security of cloud data is increasingly important. Attribute-based encryption (ABE) has been a focus of attention in recent years as an important direction for achieving cloud data encryption security due to its fine-grained access control. Although ABE itself has good access control capabilities, most existing solutions still lack effective mechanisms for comparing numerical attributes to perform access control, and the access policy in the ciphertext may expose sensitive information.
In this thesis, we present a generic construction for a range comparable attribute-based encryption scheme with partially hidden access policies. This construction utilizes 0-Encoding and 1-Encoding techniques to enable numerical attribute comparisons in access policies, but also considers partially hidden access policies in the mechanism to address the issue of sensitive information leakage from access policies. We also propose a method to address privacy concerns that may arise when combining partially hidden access policies with the transformation of attributes using 0-Encoding and 1-Encoding techniques.
Furthermore, in comparison to other comparable attribute-based encryption schemes, which only allow access policies to specify permissible attribute ranges, our approach takes it a step further by enabling attributes of users to encompass ranges concept, rather than being limited to single elements. This enhancement significantly enhances the flexibility of the proposed scheme, making it more adaptable to a wider range of scenarios
Genome-wide identification and classification of the truncated Hsp70 (Hsp70T) in plant kingdom
No full textThe heat shock protein 70 (Hsp70) is a vital chaperone protein involved in facilitating proper protein folding. In the plant kingdom, four distinct types of Hsp70 have been identified: cytosolic Hsp70 (cyHsp70), ER Hsp70 (Bip), mitochondrial Hsp70 (mtHsp70) , and chloroplast Hsp70 (cpHsp70). All types of Hsp70 contain a conserved C-terminal motif that enables co-chaperone interactions or protein retention. The truncated forms lacking this motif, known as truncated Hsp70 (Hsp70T), have an unknown function in plants. To explore the distribution and protein features of Hsp70T, we performed a comprehensive genome-wide analysis using data from 192 sequenced plant and algal genomes. Our findings revealed that Hsp70T is present in all four types of Hsp70, with truncated cyHsp70 (cyHsp70T) being the most prevalent and abundant. The cyHsp70T sequences are classified into two lineages: T1 and T2. T1 members exhibit higher similarity to cyHsp70s compared to T2 members. Notably, T1 sequences display significant diversity, leading to the establishment of four subtypes. T1 is present in approximately half of the investigated algal and plant species, with varying numbers ranging from one to 48, while T2 is found in almost all plants, typically with fewer than five copies. Furthermore, T1 has expanded into tandem gene clusters known as T1 clusters in 26 T1-rich plants. These T1 clusters show a close evolutionary relationship within species from the same family. Intriguingly, the T1 clusters of Lauraceae species share a close evolutionary relationship with those of basal angiosperms, indicating a common ancestral T1 cluster. Transcript level analysis revealed that T1 is expressed at low levels, while T2 shows specific expression in seeds and under heat stress, suggesting involvement of T2 in seed development and stress adaption