Repository logo
  • English
  • ไทย
Log In
New user? Click here to register. Have you forgotten your password?
Communities & Collections
All of DSpace
punsarn.menu.mahidol_journal
punsarn.menu.statistics
punsarn.menu.about.us
punsarn.menu.customer.feedback
punsarn.menu.deposit
  1. Home

Browsing by Author "Samanchuen T."

Filter results by typing the first few letters
Now showing 1 - 20 of 21
  • Results Per Page
  • Sort Options
  • No Thumbnail Available
    ItemMetadata only
    A Flexible Water Monitoring System for Pond Aquaculture
    (2022-01-01) Wannee N.; Samanchuen T.; Mahidol University
    Water quality is the most important factor for pond aquaculture. Farmers must be able to control a water quality to be within the appropriate range for each kind of aquatic animal, which needs both skill and knowledge. This work proposes a flexible water quality monitoring system for pond aquaculture, which can be configured to fit with a particular aquatic animal. Three types of sensors are applied to measure the water quality including pH, temperature, and dissolved oxygen sensors with 2 types of wireless connection, i.e., WiFi and NB-IoT. A microcontroller is used to process these sensing signals and send data to a cloud server. The system monitors the water quality all the time. If the water quality is not within the controlled range, the alert message will be sent to the farmers. The dashboard is also provided for real-time system monitoring. Systems can be selected as three kinds of aquatic animals including white shrimp, white sea bass, and tilapia. The controlled range will be changed according to the type of aquatic animal that the user selected. As the result, we found that the proposed system can perform as designed and the accuracies of the prototype are 98.5%, 97.6%, and 95.1% for pH, temperature, and dissolved oxygen, respectively, when compared with the standard tools. The proposed system can support both connections, i.e., WiFi and NB-IoT.
  • No Thumbnail Available
    ItemMetadata only
    A Machine Learning Approach to Identifying Suspicious Tax Evasion Behavior in Public Financial Data
    (2023-01-01) Visitpanya N.; Samanchuen T.; Mahidol University
    The protection of sensitive tax data poses a challenge for data processors seeking to use it freely, particularly in the realm of financial and tax data. Detecting suspicious behavior related to tax evasion in this area requires sophisticated techniques. This paper presents a study on detecting tax evasion behavior using publicly available financial data. The data was obtained from a public source, and machine learning was applied to identify suspicious behavior based on criteria derived from a literature review. Our study yielded a precision of over 97 % using selected supervised models. Future work involves exploring behavior using unsupervised methods, increasing the number of instances using data synthesis techniques, and comparing results with enterprises that are not registered in a stock market as long as the data is legally available. This research provides valuable insights into identifying suspicious tax evasion behavior and can inform the development of improved tax evasion detection systems in the future.
  • No Thumbnail Available
    ItemMetadata only
    A precision medicine approach to personalized prescribing using genetic and nongenetic factors for clinical decision-making
    (2023-10-01) Jamrat S.; Sukasem C.; Sratthaphut L.; Hongkaew Y.; Samanchuen T.; Mahidol University
    Screening potential drug–drug interactions, drug–gene interactions, contraindications, and other factors is crucial in clinical practice. However, implementing these screening concepts in real-world settings poses challenges. This work proposes an approach towards precision medicine that combines genetic and nongenetic factors to facilitate clinical decision-making. The approach focuses on raising the performance of four potential interaction screenings in the prescribing process, including drug–drug interactions, drug–gene interactions, drug–herb interactions, drug–social lifestyle interactions, and two potential considerations for patients with liver or renal impairment. The work describes the design of a curated knowledge-based model called the knowledge model for potential interaction and consideration screening, the screening logic for both the detection module and inference module, and the personalized prescribing report. Three case studies have demonstrated the proof-of-concept and effectiveness of this approach. The proposed approach aims to reduce decision-making processes for healthcare professionals, reduce medication-related harm, and enhance treatment effectiveness. Additionally, the recommendation with a semantic network is suggested to assist in risk–benefit analysis when health professionals plan therapeutic interventions with new medicines that have insufficient evidence to establish explicit recommendations. This approach offers a promising solution to implementing precision medicine in clinical practice.
  • No Thumbnail Available
    ItemMetadata only
    An Innovation Framework of Medical Organic Cannabis Traceability in Digital Supply Chain
    (2022-12-01) Pookkaman W.; Samanchuen T.; Mahidol University
    Cannabis is increasingly accepted by medical organizations for medicinal and research purposes. A traceability system is required for monitoring and controlling the use of cannabis. This work aimed to investigate the relationship between critical success factors (CSFs) for creating the innovation framework that affects the implementation of the cannabis traceability system. These factors are identified based on the digital supply chain by structural modeling. Additionally, the issue of patients’ safety is a crucial factor that needs to be considered. Total Interpretive Structural Modeling (TISM) and fuzzy MICMAC analysis techniques were applied to investigate the relationship between CSFs. Finally, the 14 CSFs affecting cannabis traceability systems were determined. The research results discovered that system reliability is the most significant key factor with the greatest influence and contribution to the achievement. The five CSFs based on TISM comprise harmonized strategic orientation, inter-organization collaboration, standardization, government regulations, and customer awareness. According to the conclusive dependence–driving power diagram, they have the most driving power in the implementation and operation traceability system. The innovation framework helps to establish the traceability system in a way that works well. This will build trust in the supply chain and prevent abuse and substandard products.
  • No Thumbnail Available
    ItemMetadata only
    Comparative Evaluation of Large Language Models for Automated Text Classification and Data Labeling in Drug Recommendation
    (2025-01-01) Rattanachayabun K.; Jamrat S.; Samanchuen T.; Rattanachayabun K.; Mahidol University
    Data labeling remains one of the most laborintensive stages in artificial intelligence (AI) development, particularly in natural language processing. This study investigates the potential of large language models (LLMs) to automate text classification and reduce the burden of manual annotation. Four state-of-the-art LLMs-GPT-4o, Llama 3.3, Llama 3.1, and Gemma 2-were evaluated on a drug recommendation dataset using three prompting strategies: zero-shot, one-shot, and few-shot learning. Model performance was assessed using F1-scores across therapeutic consideration, use as directed, and contraindication categories. The results indicate that incorporating even a single labeled example substantially improves classification accuracy compared with zero-shot prompting, while the performance gain from one-shot to few-shot prompting is marginal. Among the tested models, Llama 3.3 achieved the most consistent results, whereas Gemma 2 demonstrated strong zeroshot generalization and GPT-4o provided stable cross-strategy performance. The findings highlight the feasibility of employing LLMs for automated data labeling and underscore the efficiency of one-shot prompting as a practical balance between accuracy and computational cost.
  • No Thumbnail Available
    ItemMetadata only
    Correction: Development of Reference Process Model and Reference Architecture for Pharmaceutical Cold Chain (Sustainability, (2023), 15, 5, (3935), 10.3390/su15053935)
    (2023-05-01) Latha D.S.; Samanchuen T.; Mahidol University
    The authors would like to make the following corrections about the published paper [1]. The changes are as follows: Authors would like to change the caption of the Figure 4, so we need to update the caption of original Figure 4:Figure 4. The final design of RPM for the pharmaceutical cold chain process where CC denotes cold chain.With:Figure 4. The final design of RPM for the pharmaceutical cold chain process adapted from [4]. Authors would like to add Acknowledgments:Acknowledgments: The authors would like to deeply acknowledge Peik Bremer, Technical University of Applied Sciences Wuerzburg-Schweinfurt, for his research contribution and valuable support to us. Authors would like to change the caption in Figure A3 in Appendix B, so we need to update the caption of original Figure A3:Figure A3. Developed RPM for pharmaceutical cold chain (a) Pack process (b) Load process (c) Transport process (d) Unload process (e) Unpack process (f) Transfer process (g) Keep in Storage process (h) Inspect process (i) Audit process.With:Figure A3. Developed RPM for pharmaceutical cold chain (a) Pack process (b) Load process (c) Transport process (d) Unload process (e) Unpack process (f) Transfer process (g) Keep in Storage process (h) Inspect process (i) Audit process. Note: Figure A3b,d,e,h reproduced with permission from Bremer, P. [4]. Authors would like to change the caption of the Figure 4, so we need to update the caption of original Figure 4: Figure 4. The final design of RPM for the pharmaceutical cold chain process where CC denotes cold chain. With: Figure 4. The final design of RPM for the pharmaceutical cold chain process adapted from [4]. Authors would like to add Acknowledgments: Acknowledgments: The authors would like to deeply acknowledge Peik Bremer, Technical University of Applied Sciences Wuerzburg-Schweinfurt, for his research contribution and valuable support to us. Authors would like to change the caption in Figure A3 in Appendix B, so we need to update the caption of original Figure A3: Figure A3. Developed RPM for pharmaceutical cold chain (a) Pack process (b) Load process (c) Transport process (d) Unload process (e) Unpack process (f) Transfer process (g) Keep in Storage process (h) Inspect process (i) Audit process. With: Figure A3. Developed RPM for pharmaceutical cold chain (a) Pack process (b) Load process (c) Transport process (d) Unload process (e) Unpack process (f) Transfer process (g) Keep in Storage process (h) Inspect process (i) Audit process. Note: Figure A3b,d,e,h reproduced with permission from Bremer, P. [4]. The authors apologize for any inconvenience caused and state that the scientific conclusions are unaffected. The original publication has also been updated.
  • No Thumbnail Available
    ItemMetadata only
    Design of A Curriculum for Online Competency-Based Training
    (2022-01-01) Hassadeepong J.; Samanchuen T.; Mahidol University
    The concept of making occupational standards as a framework to develop a workforce has been adopted in many countries. Each occupation requires certain skills and capabilities necessary to pursue that profession. However, occupational standards are only a description of those desired skills. A curriculum is an important part of developing an effective and productive workforce. Therefore, this research presents a method/concept for designing a competency-based curriculum using the principles of Outcome-Based Education (OBE) and use the National Occupational Standards (NOS) as a source of basic information for development by converting competency units into the curriculum, The goal is to be a competency-based curriculum for online competency-based training because due to the current situation, learning is more likely to be online platforms. The occupational standard of data analyst is used to demonstrate the process of this work. The conceptual framework was assessed using the Delphi technique by 10 experts. The results show that the proposed framework is suitable for developing a competency-based curriculum. As a result of this method, the curriculum is complete and consistent with occupational standards.
  • No Thumbnail Available
    ItemMetadata only
    Developing a comprehensive and strategic sustainability framework for the evolving sharing economy
    (2026-06-01) Khongkhoom P.; Samanchuen T.; Khongkhoom P.; Mahidol University
    This study develops and validates a comprehensive sustainability framework for sharing-economy businesses (SEBs), positioning sustainability as an integrated managerial agenda rather than a set of isolated initiatives. Drawing on an extensive review of sustainability and sharing-economy literature, the framework consolidates environmental, social, economic, and technological dimensions and embeds them across three organizational levels: top management, middle management, and operational management. A structured three-phase research process was employed, comprising the identification and alignment of sustainability factors with sharing-economy business operations, the development of organization-aligned recommended solutions, and the integration of these solutions into a multi-level framework grounded in the sustainable project life-cycle perspective. The framework and proposed solutions were validated using the Delphi technique involving experts from academia, industry, government, platform operators, asset owners, and users. Consensus analysis based on median values, standard deviation, and interquartile range demonstrates strong convergence and stability of expert judgments across all validated items. The findings indicate that effective sustainability implementation in SEBs depends on clear differentiation of managerial roles, coordinated decision-making across organizational levels, and robust digital-platform capabilities that support trust, transparency, and governance. By bridging the gap between strategic intent and operational execution, the proposed framework provides an empirically grounded framework for platform-based organizations seeking to embed sustainability into governance structures and daily operations across diverse sharing-economy contexts.
  • No Thumbnail Available
    ItemMetadata only
    Developing a Comprehensive Virtual Learning Management Framework: Key Factors and Strategies for Success
    (2024-01-01) Chuaphun P.; Chujai L.; Samanchuen T.; Chuaphun P.; Mahidol University
    —This research addresses the pressing need for effective virtual learning management frameworks (VLMFs) in traditional educational settings. The study explores the key factors influencing virtual learning success and develops a comprehensive VLMF. The framework integrates technology and management as fundamental drivers of success, emphasizing their interplay in creating a conducive virtual learning environment. The research objectives are summarizing key success factors, exploring the VLMF, and developing a conceptual framework integrating technology and management. The study’s significance lies in its contribution to the understanding and implementing virtual learning, highlighting the transformative potential of integrating technology and management principles. The findings provide valuable insights into the interconnected factors crucial for virtual learning success, underscoring the importance of a well-coordinated approach in navigating the complexities of virtual education. The study’s results have implications for teaching and learning, emphasizing the need for a structured approach to enhance learning outcomes in virtual education.
  • No Thumbnail Available
    ItemMetadata only
    Development of Reference Process Model and Reference Architecture for Pharmaceutical Cold Chain
    (2023-03-01) Latha D.S.; Samanchuen T.; Mahidol University
    Although the competence level in pharmaceutical cold chain logistics is adequate in the current healthcare sector, the future will be more unpredictable, disruptive, and chaotic than the world of today because of rapid technological changes, as well as social disruption. This work introduces and evaluates a new cold chain structure based on the enhanced reference process model (RPM) and reference architecture (RA) for the pharmaceutical cold chain competence design. The modified Delphi technique is used to design a systematic decision context to evaluate the improved RPM and RA and produce a competence design by obtaining agreement from a panel of experts. Ten experts in cold chain logistics took part in the modified Delphi assessment process to describe the model and architecture for an evaluation lead. They defined it as the assessment associated with the professional related to cold chain accreditation. Ten participants from six countries were asked questions regarding their expertise, abilities, and opinions in the first round, and their comments were collected. In the second and third rounds, comments and consensus were collected, which were set at 80% for RPM and RA. This work proposes an innovative scheme for developing occupational standards based on the RPM and RA, as opposed to the traditional method of employing functional analysis to establish occupational standards. This work can be further expanded to develop professional competencies in the pharmaceutical cold chain.
  • No Thumbnail Available
    ItemMetadata only
    Domain-Specific Named Entity Recognition in Hotel Reviews Using Large Language Models
    (2025-01-01) Tiebrat K.; Chuckpaiwong R.; Samanchuen T.; Tiebrat K.; Mahidol University
    Extracting meaningful insights from large volumes of hotel review text remains a key challenge in natural language processing. Traditional methods often rely on manual feature design or limited rule-based systems. This study introduces an automated workflow that uses large language models (LLMs) to transform unstructured hotel reviews into structured, analytical data. The process integrates predefined prompts with the review texts to guide the model in identifying entities, classifying them into relevant aspects, and determining sentiment polarity. The approach effectively captures contextual information and produces highly interpretable, structured outputs. By leveraging prompt engineering and LLM inference, it reduces human intervention while maintaining analytical depth. The proposed workflow demonstrates the potential of LLM-driven automation to streamline sentiment and aspect extraction from usergenerated content, enabling scalable and data-driven insights in hospitality and related domains.
  • No Thumbnail Available
    ItemMetadata only
    Enhancing Cyberbullying Detection on Social Media Using Transformer Models
    (2024-01-01) Tapaopong W.; Charoenphon A.; Raksasri J.; Samanchuen T.; Tapaopong W.; Mahidol University
    —The proliferation of cyberbullying on social media platforms has led to devastating physical, emotional, and mental consequences for victims, underscoring the need for effective identification and analysis methods. This study leverages Natural Language Processing (NLP) techniques, specifically Transformer models, to enhance cyberbullying detection in harmful social media messages. By employing transfer learning and fine-tuning strategies, we compared the effectiveness of category prediction using five prominent Transformer models: BERT, RoBERTa, ALBERT, DistilBERT, and ConvBERT. Our results demonstrate that DistilBERT outperforms other models, achieving an accuracy of 94.36%, precision of 93.83%, and recall of 93.91% in the 5-Class categorization of cyberbullying severity. This study highlights the significance of automated cyberbullying detection and underscores the potential of NLP techniques in combating the adverse impacts of cyberbullying. The findings of this research have important implications for the development of effective cyberbullying detection systems, which can help mitigate the harmful effects of online harassment. By leveraging the capabilities of DistilBERT, we can create more accurate and efficient systems for detecting and preventing cyberbullying, ultimately contributing to a safer online environment.
  • No Thumbnail Available
    ItemMetadata only
    Evaluating Wireless Power Transfer Technologies for Electric Vehicles: Efficiency and Practical Implementation of Inductive, Capacitive, and Hybrid Systems
    (2025-01-01) Singhavilai T.; Tippayachai J.; Jirasereeamornkul K.; Ekkaravarodome C.; Samanchuen T.; Singhavilai T.; Mahidol University
    This study evaluated wireless power transfer (WPT) technologies for electric vehicles (EVs), focusing on inductive (IPT), capacitive (CPT), and hybrid (HPT) systems. IPT utilizes resonant magnetic fields, CPT employs resonant electric fields, and HPT combines both methods to optimize the use of electromagnetic fields and electronic components. Pilot experiments were conducted using WPT standards and the relevant literature to investigate the efficiency and practical implementation of these WPT technologies. The evaluation included measuring system efficiencies with multimeters, assessing input and output waveform smoothness using oscilloscopes, detecting power losses through thermal scans, and monitoring electromagnetic field (EMF) exposure with EMF detectors. The results demonstrated that IPT achieved higher efficiency, smoother waveforms, and lower EMF exposure than CPT and HPT at lower frequencies. Moreover, IPT has a more straightforward circuit design owing to the lack of high-frequency components, further enhancing its practicality. The study also examined the effects of ground clearance and misalignment on WPT performance, and addressed safety concerns and potential solutions for all three types of WPT systems.
  • No Thumbnail Available
    ItemMetadata only
    Exploring success factors and relationships in virtual learning using ISM and fuzzy MICMAC analysis
    (2024-04-15) Chuaphun P.; Samanchuen T.; Chuaphun P.; Mahidol University
    The COVID-19 pandemic's consequences have led to a global change in educational settings towards online learning. The utilization of virtual learning (VL) has increased significantly. This study aimed to extract the success factors of VL and also examine the relationships among them. The research method involves examining factors identified in the literature review and seeking confirmation from experts using the Content Validity Index (CVI) method. Ten success factors are extracted and confirmed, including Technological, Management, Learning Capability, Pedagogical, Ethical, Resource Support, Interface Design, Evaluation, Institutional, and Study Environment. Based on the Interpretive Structural Model (ISM) method and the fuzzy matrix of cross-impact multiplications applied to classification (MICMAC), which divides the factors into five levels, the relationship between these factors is examined. Level I emphasizes the importance of evaluation mechanisms. Level II stresses integrating pedagogical, ethical, resource support, and institutional aspects. Level III highlights the alignment of learner capabilities with platform interfaces. Level IV underscores the significance of the learning environment. Lastly, Level V emphasizes the interplay between technology and management in VL's expansion. The findings of this study can be developed and customized through collaboration among instructors, learners, and institutions. Moreover, the findings from correlating success factors can be applied in practical learning experiments or utilized to develop efficient modeling manuals.
  • No Thumbnail Available
    ItemMetadata only
    Factors Influencing Behavior Intention in Digital Investment Services of Mutual Fund Distributors Adoption in Thailand
    (2023-02-01) Kasemharuethaisuk H.; Samanchuen T.; Mahidol University
    There are various types of mutual fund distributors in Thailand that utilize technology to provide investment services to individual investors. These services can be accessed through mobile or internet banking, allowing investors to make transactions and invest in mutual funds at their convenience. This work aims to identify the factors that influence individual investors in Thailand to use digital investment services offered by brokerages. We have developed a conceptual model based on the Technology Acceptance Model (TAM2) and relevant literature on fintech and financial behavior, comprising seven variables and six hypotheses. Our research method involves a questionnaire survey of Thai investors and the use of partial least square structural equation modeling (PLS-SEM) for data analysis. The results show that individual investors’ intention to use digital investment services is significantly impacted by their perception of the usefulness of these services. Additionally, this intention is also influenced by other variables such as convenience, trust, and subjective norm, but not by perceived ease of use. This may be because most of our study’s participants are tech-savvy. Our findings provide insight into the perspectives and perceptions of Thai individual investors who have experiences of mutual fund investment.
  • No Thumbnail Available
    ItemMetadata only
    Implementation of TISM and fuzzy MICMAC analysis in Jupyter notebook with Python
    (2024-01-01) Pookkaman W.; Samanchuen T.; Sirisawat S.; Boonkong K.; Pookkaman W.; Mahidol University
    Total Interpretive Structural Modeling (TISM) and fuzzy Matrixed Impacts Croises Multiplication Appliquee a a Classement (fuzzy MICMAC) are factor analysis techniques to evaluate the relationships among the factors on one interest topic. More works have started to use these techniques as the primary analysis tool. This paper introduces a new implementation of TISM and fuzzy MICMAC analysis using Jupyter Notebook and Python. The motivation behind this implementation is the limited availability of suitable software tools for conducting this type of analysis. The methodology used to implement the function is detailed in the paper. The TISM analysis function is implemented using an iterative algorithm approach, while the fuzzy MICMAC function is implemented using a matrix approach. The function allows for user input of criteria and factors, and outputs interpretive structural modeling and cross-impact matrix calculations, as well as a stability graph for fuzzy MICMAC analysis. To validate the implementation, the results generated by the function were compared to those obtained in previous research. The comparison showed that our implementation provides accurate results that align with those generated by the previous research. Furthermore, the function was found to be easy to use and allowed researchers to input their data and obtain results quickly.
  • No Thumbnail Available
    ItemMetadata only
    Improving Course Registration Efficiency Through LINE Chatbot with LLM Integration
    (2025-01-01) On-Uean K.; Chaikot A.; Samanchuen T.; Pookkaman W.; On-Uean K.; Mahidol University
    This study addresses inefficiencies in course registration at a private training center that used Google Forms. The process resulted in redundant data entry, a 25% error rate, and no real-time status tracking. The study pursued three objectives 1) developing an Automated Question Answering System (AQAS) using LLM technology 2) implementing an Automated Course Registration System (ACRS) via LINE Messaging API and 3) minimizing data entry errors to enhance data management efficiency. System development followed the Software Development Life Cycle (SDLC) waterfall model, the system was developed through structured phases of analysis, design, implementation, testing, deployment, and maintenance. Data management utilizes MySQL database with administrator validation conducted directly through LINE. System evaluation presented improvements in all three objectives. The LLMintegrated question-answering system achieved 86% accuracy across 50 structured queries. Comparative testing with 80 participants across three courses showed that the LINE-based registration system improved registration data entry accuracy by a 26.7% relative increase and reduced processing time by 50-66 %. These improvements enhanced user interaction quality, reduced staff workload, and demonstrated the effectiveness of AI-enabled chatbots for educational administration.
  • No Thumbnail Available
    ItemMetadata only
    Predicting and Managing EV Charging Demand on Electrical Grids: A Simulation-Based Approach
    (2023-04-01) Jaruwatanachai P.; Sukamongkol Y.; Samanchuen T.; Mahidol University
    Electric vehicles (EVs) are becoming increasingly popular, and it is important for utilities to understand their charging characteristics to accurately estimate the demand on the electrical grid. In this work, we developed simulation models for different EV charging scenarios in the home sector. We used them to predict maximum demand based on the increasing penetration of EV consumers. We comprehensively reviewed the literature on EV charging technologies, battery capacity, charging situations, and the impact of EV loads. Our results suggest a method for visualizing the impact of EV charging loads by considering factors such as state of charge, arrival time, charging duration, rate of charge, maximum charging power, and involvement rate. This method can be used to model load profiles and determine the number of chargers needed to meet EV user demand. We also explored the use of a time-of-use (TOU) tariff as a demand response strategy, which encourages EV owners to charge their vehicles off-peak in order to avoid higher demand charges. Our simulation results show the effects of various charging conditions on load profiles and indicate that the current TOU price strategy can accommodate a 20% growth in EV consumers, while the alternative TOU price strategy can handle up to a 30% penetration level.
  • No Thumbnail Available
    ItemMetadata only
    Revolutionizing Pharmaceutical Cold Chain Competency Framework with Reference Process Model and Reference Architecture
    (2023-06-01) Latha D.S.; Samanchuen T.; Mahidol University
    The utilization of the reference process model (RPM) and reference architecture (RA) as disruptive information and communication technologies (ICT) in the pharmaceutical cold chain (PCC) industry has enabled the management of tasks through a model-architecture-based approach. This research presents an innovative method for competency development in the cold chain sector, leveraging RPM and RA. By introducing a comprehensive conceptual framework encompassing RPM and RA design and workflow into cold chain competency development, this study outlines the key areas for incorporating RPM and RA into the PCC field. The framework elucidates the functioning of RPM and RA concerning occupational standards (OS) and units of competencies (UOC) within the industry. The study generates a research framework for the PCC industry by systematically implementing RPM and RA using a proposed method. The primary outcomes and empirical evidence are UOC and OS derived from RPM and RA implementation and integration, substantiating the conceptual framework's validity. The research highlights the evolutionary aspects and the significance of the conceptual framework in guiding the research framework and proposes a method for competency development. Furthermore, recommendations are provided for future research endeavors.
  • No Thumbnail Available
    ItemMetadata only
    Sentiment Analysis of Thai Stock Reviews Using Transformer Models
    (2022-01-01) Harnmetta P.; Samanchuen T.; Mahidol University
    The stock market is typically affected by various factors for a long time, such as politics, economics, and finance. These are expressed through online media that people can easily access today. Moreover, in the digital era, data growth is an exponential trend, and a million records of data are generated through many online platforms over the internet. To utilize that information in time, a stock sentimental analysis system integrated with the transformer base model is proposed. This work applies the transformer base models that can break through NLP limitations from the past. Furthermore, we gather data as fundamental analysis in Thai financial content from a financial institution. However, to compare the result between embedding techniques with baseline, we use multinomial logistic regression in the form of a predictive model and apply the baseline, the term frequency-inverse document frequency (TF-IDF). Our experiment shows that WangchanBERTa and BERT can achieve high accuracy at 92.52% and 89.12%, respectively, and the baseline result is 85.03%. In conclusion, our proposed system can precisely predict stock sentiment in Thai with high accuracy.
  • «
  • 1 (current)
  • 2
  • »
footer.content.left
Certified Logo

มตช./ข้อกำหนดเผยแพร่ 2001-2567 คลังสารสนเทศดิจิทัล - ข้อกำหนด

TCAS/TPAS 2001:2024 DIGITAL REPOSITORIES - REQUIREMENTS

  • Privacy policy
  • End User Agreement