Client Satisfaction Analysis for Delivery of Services with Natural Language Processing and Decision Support System

Client Satisfaction Analysis for Delivery of Services with Natural Language Processing and Decision Support System

Jemar Almaceda Banawa, Mervin Jommel Tibay De Jesus

Laguna State Polytechnic Univerity Siniloan Laguna, Philippines

DOI: https://dx.doi.org/10.47772/IJRISS.2025.905000283

Received: 07 May 2025; Accepted: 09 May 2025; Published: 11 June 2025

ABSTRACT

Traditional client satisfaction surveys in state universities and government agencies often suffer from inefficiencies in data collection, analysis, and visualization. Manual processing leads to delays, inaccuracies, and limited actionable insights, hindering effective decision-making. This study aims to improve client satisfaction surveys by developing a web-based application that integrates Natural Language Processing (NLP) and Decision Support Systems (DSS). The goal is to automate the collection, analysis, and visualization of feedback, enhancing data-driven decision-making and service improvement. A developmental and descriptive research design was used, with data collected from university employees and clients involved in service delivery and surveys. Stratified sampling ensured diverse representation from faculty and students. The system was developed using Agile methodologies, allowing for iterative improvements based on user feedback. NLP was applied to analyze open-ended responses, while DSS was used to generate actionable insights. The system reduced survey processing delays by automating data analysis and visualization. NLP sentiment analysis improved the accuracy of open-ended feedback. Real-time insights were provided through interactive dashboards, aligning with the Anti-Red Tape Authority’s (ARTA) goal of improving government service efficiency. The web-based application effectively solved inefficiencies in traditional survey methods by automating key processes. Integrating NLP and DSS improved data accuracy, reduced delays, and enhanced service delivery in government institutions. State universities and government agencies should adopt this approach to enhance the efficiency and transparency of client satisfaction surveys, with further research exploring its application in other government processes.

Keywords: Client Satisfaction Surveys, Natural Language Processing, Decision Support Systems, Sentiment Analysis, Government Efficiency.

INTRODUCTION

Client satisfaction is crucial for evaluating the quality of services in state universities and local government units. Traditionally, feedback is collected through paper surveys, which often result in slow responses, missing data, and added administrative work. While many organizations are turning to digital tools, many public institutions in the Philippines still rely on manual methods, which limit timely decision-making and data analysis [1].

Recent developments in technologies like Natural Language Processing and Decision Support Systems offer promising solutions to improve how client feedback is collected and analyzed. NLP helps automatically sort feedback into positive and negative responses, while DSS can help make sense of the data to guide improvements[2]. However, challenges such as lack of standardization, limited budgets, and varying technical skills still prevent many institutions from fully adopting these digital solutions [3].

The establishment of the Anti-Red Tape Authority through Republic Act 11032 highlights the importance of improving government services and measuring client satisfaction effectively [4]. Despite efforts to go digital, many state universities and local government unit still struggle to implement effective feedback systems, often relying on outdated methods [5].

This study proposes using NLP and DSS to develop a digital system for collecting and analyzing client satisfaction surveys. The goal is to replace manual methods with a faster, more accurate process that provides real-time insights. By applying machine learning tools, such as sentiment analysis and decision tree models, this system will help decision-makers in public institutions improve services based on reliable and timely data [6].

Research Objectives

The primary aim of this study is to develop a website application for client satisfaction analysis and service delivery that uses natural language processing and a decision-support system.

Specific Objectives

1. Design and develop a website application that facilitates collection of clients or customer satisfaction data.
2. Incorporate Natural Language Processing (NLP) to analyze and categorize qualitative survey responses for deeper insights into client feedback.
3. Implement Decision Support Systems (DSS) to generate actionable insights from survey data and aid in data-driven decision-making processes.
4. Utilize data analytics to visualize and interpret survey data, enabling effective monitoring of service performance and trends.
5. Assess the effectiveness and usability of the developed system based on ISO 25010 quality standards.
6. Evaluate the user’s level of acceptability based on system useability scale.

Fig 1. Customer Client Satisfaction Survey System Conceptual Framework

As seen in Fig 1, the study utilizes the Input-Process-Output (IPO) model to develop a client satisfaction analysis system integrating Natural Language Processing and Decision Support Systems. The system is designed to optimize the client satisfaction survey process for State Universities and Colleges, enabling data-driven decision-making to improve service quality.

In the input phase, the system allows evaluators to provide feedback through a structured Customer Satisfaction Survey (CCSS) form. This form includes key sections for evaluating service quality, adherence to government standards, and capturing open-ended feedback. Registered users benefit from an auto-fill feature to streamline the feedback process.

The process phase involves multiple stages of data preparation. Textual feedback is preprocessed using techniques like stop word removal and lemmatization. NLP tools such as syntax parsing and semantic analysis are applied to extract meaningful insights. Sentiment analysis is performed using the Iterative Dichotomiser 3 (ID3) decision tree algorithm, categorizing the feedback as positive or negative based on predefined sentiment scores. Additionally, the system employs a dual DSS approach: a Data-Driven DSS to analyze historical data and a Model-Driven DSS to process real-time data, allowing the system to dynamically adapt to new feedback.

At the core of the system is a Large Language Model (LLM), which processes the data through stages such as filtering, embedding, and fine-tuning, and utilizes Reinforcement Learning with Human Feedback (RLHF) for continuous improvement. The LLM enhances the decision-making process by refining the data analysis and producing actionable insights.

In the output phase, the system provides visualized insights and reports tailored to different administrative roles. Administrator 1 (Office/Unit) receives feedback and performance ratings for their specific unit, enabling targeted improvements. Administrator 2 (MIS Office) accesses aggregated feedback from multiple units and generates detailed reports for higher management. The Super Administrator (Campus Director) has full access to all data, facilitating organization-wide analysis and decision-making.

This IPO-based framework enables effective feedback analysis and reporting, empowering universities to enhance service quality and make informed decisions based on comprehensive client feedback.

Review Of Related Literature

Customer/Client Feedback in Government Agencies

Customer feedback plays a vital role in enhancing service delivery in government agencies and state universities by identifying areas for improvement and fostering transparency and accountability. This process allows institutions to refine their services, ensuring they meet public expectations and maintain public trust [7]. Gathering feedback also helps public institutions develop citizen-centered services, ensuring that services are tailored to community needs, thereby improving overall governance and engagement[8]. However, many agencies face challenges with outdated feedback methods and a lack of adequate staff training, which undermine the quality of feedback [9].

Digitalization Initiatives of the Philippine Government

The Philippine government has undertaken digitalization initiatives to improve service delivery and enhance transparency, such as the Philippine Digital Strategy and the e-Government Master Plan. These efforts aim to streamline processes and make government services more accessible [10]. Despite these positive steps, challenges remain, particularly in rural areas where digital infrastructure is lacking, and in government agencies where technical expertise and training are limited [11]. Addressing these issues is crucial for ensuring that digitalization benefits all sectors equally and fosters public trust through better access to services [12].

ARTA Memorandum Circular No. 2023-05

The Anti-Red Tape Authority Memorandum Circular No. 2023-05 mandates client satisfaction surveys for government agencies and state universities to enhance service delivery. This initiative promotes transparency and accountability by ensuring that public institutions gather and act on client feedback to improve their services [13]. However, challenges such as resource limitations, insufficient survey design expertise, and data privacy concerns hinder the effectiveness of this initiative [14]. Addressing these barriers is essential for maximizing the potential of client feedback and improving public service [15].

Natural Language Processing

NLP enhances the analysis of client satisfaction surveys by extracting meaningful insights from qualitative data, such as open-ended responses. By categorizing feedback and identifying emerging concerns, NLP helps public institutions make more informed decisions [16]. However, challenges such as data quality, lack of technical expertise, and language nuances hinder the practical application of NLP in many government agencies and universities [17]. Despite these obstacles, the integration of NLP offers significant benefits in understanding client sentiments and improving service delivery [18].

Comparison of Decision Tree Algorithms

Among decision tree algorithms, ID3 (Iterative Dichotomiser 3) is considered the best for sentiment analysis and text classification tasks due to its simplicity, interpretability, and solid performance in balancing precision and recall. Unlike more complex algorithms like Random Forest or CART, ID3 produces clear, understandable decision trees that are ideal for transparency and real-time applications (Vaswani et al., 2017). Although Random Forest offers higher accuracy, it lacks interpretability, making ID3 a more suitable choice for applications that require ease of explanation and debugging as shown in table 1.

DSS for Customer/Client Satisfaction Surveys

Decision Support Systems are crucial for analyzing customer/client satisfaction surveys in government agencies and state universities. Data-driven DSS can process historical survey data, identifying patterns and gaps in service delivery, while model-driven DSS uses real-time data to predict client reactions to service changes [19]. Integrating Natural Language Processing (NLP) with DSS enhances the analysis of open-ended responses, providing a comprehensive view of client satisfaction and enabling informed, proactive decision-making [20].

Descriptive Data Analysis

Descriptive data analysis is vital for summarizing large datasets from client satisfaction surveys into meaningful insights. It helps organizations understand key issues affecting client satisfaction, such as service delivery gaps, and supports decision-making [21]. However, many organizations struggle with this analysis due to a lack of technical expertise and limited data collection methods [22]. Integrating advanced tools like NLP can help institutions process qualitative feedback effectively, providing a fuller picture of client needs and improving the accuracy of service adjustments [23].

Large Language Model (LLM)

LLMs offer powerful tools for analyzing client satisfaction surveys by processing large volumes of structured and unstructured data. LLMs can classify sentiments, extract key themes, and automate feedback analysis, streamlining the process and improving service quality [24]. However, implementing LLMs in government institutions is challenging due to the computational resources required and concerns over data privacy [25]. Despite these challenges, LLMs hold significant potential for enhancing feedback analysis, providing deeper insights into client satisfaction [26]

METHODOLOGY

This study utilized a combination of developmental and descriptive research designs to create an automated client satisfaction analysis system that incorporates Natural Language Processing and a Decision Support System. The developmental research design followed a systematic process involving system planning, prototype development, expert evaluation, and iterative refinements based on stakeholder feedback. This approach ensured continuous improvements in system functionality and usability, allowing for adaptive modifications throughout the development cycle [27]. It facilitated advancements in automated client feedback analysis and enhanced decision-making in service delivery.

The descriptive research design provided a framework to examine current client satisfaction survey processes, identifying challenges in data collection, interpretation, and decision-making. It involved gathering both qualitative and quantitative data from existing methodologies, analyzing feedback trends, and evaluating the limitations of manual evaluation processes [28]. Additionally, this design assessed the proposed system’s ability to process natural language responses, extract meaningful insights, and generate data-driven recommendations through NLP and DSS, while comparing its accuracy and efficiency to traditional methods [29]. By integrating both research methodologies, this study addressed existing challenges in client satisfaction analysis and contributed to the development of an innovative automated feedback processing system. The study’s findings aim to improve decision-making and service delivery in government agencies, facilitating more efficient and effective public service.

Fig 2. Agile Methodology

As shown in Fig 2, the study followed the Agile software development model to create a client satisfaction analysis system that integrates Natural Language Processing (NLP) and a Decision Support System (DSS). Agile was chosen for its iterative and incremental development approach, allowing continuous feedback and adaptation based on stakeholder input. This methodology ensures the system evolves to meet functional and user requirements effectively, with feedback loops for refinements during testing phases.

The development process was divided into key phases: planning, design, development, testing, deployment, and feedback. The planning phase focused on studying current manual processes at a university in Laguna and aligning the system with regulatory standards, particularly the ARTA guidelines. In the design phase, system architecture, user interface, and functional components were developed, with stakeholder feedback ensuring the design met user needs. The development phase translated design specifications into functional code using technologies like PHP (Hypertext Preprocessor), a widely-used scripting language for web development, MySQL (My Structured Query Language), a relational database management system for storing and managing data, CSS (Cascading Style Sheets), a stylesheet language for designing the layout and presentation of web pages, HTML (Hypertext Markup Language), the standard language used to create the structure and content of web pages, and JavaScript, a programming language that enables interactive web pages. These technologies were employed to ensure the system’s adaptability, reliability, and overall functionality.

Testing was an ongoing process, including unit, integration, and user acceptance testing to ensure the system met quality standards. Following successful testing, the deployment phase focused on preparing the application for release, including user training and post-release support. The feedback phase, which gathers insights from users, ensures continuous improvements and updates to the system.

By adopting the Agile model, this study enabled the development of a responsive, user-centered system that enhances client satisfaction survey processes in government agencies, ensuring better service delivery and informed decision-making.

RESULT AND DISCUSSION

This Result and Discussion discusses the findings from the research, with a focus on evaluating the integration of Natural Language Processing and Decision Support Systems in enhancing client satisfaction analysis. The analysis follows a systematic approach, aligning with the research objectives, and examines each component’s effectiveness in addressing the study’s research problems. Key insights are drawn from the performance, usability, and functionality tests, with recommendations for further improvements and potential applications.

Design and develop a website application that facilitates collection of clients or customer satisfaction data.

Fig 3. User Interface (UI) and User Experience (UX) prototype

A key outcome of the research was the development of a User Interface (UI) and User Experience (UX) prototype that enhances user interaction. Utilizing Microsoft PowerPoint for visualization, this prototype ensures a streamlined design that simplifies the feedback collection process. As emphasized by[30] and [31], prototyping at early stages improves user satisfaction by addressing usability issues before the final implementation. The UI/UX design serves as a foundational guide for the system’s full development, ensuring user-friendly access to the feedback collection tools as depicts in Fig 3.

Fig 4. Feedback Submission and Processing

As illustrated in Fig 4, the feedback submission interface, which includes both client and administrator access points, has been designed for ease of use. The client interface allows seamless feedback submission through structured forms, with built-in validation to ensure completeness and accuracy. The backend processing system ensures secure data storage and adheres to best practices in database management, including SQL injection prevention and efficient data retrieval through optimized queries [32].

Incorporate Natural Language Processing (NLP) to analyze and categorize qualitative survey responses for deeper insights into client feedback.

Fig 5. Natural Language Processing (NLP) Integration

A crucial component of the system is its NLP functionality, which processes open-ended client feedback to extract meaningful insights. The integration of NLP streamlines the analysis of unstructured textual data, reducing manual effort and improving consistency across feedback datasets. Text preprocessing techniques such as tokenization, stop word removal, and stemming are employed to prepare raw text for sentiment analysis. The Decision Tree-based sentiment analysis model, trained on a labeled dataset, classifies feedback as positive or negative, with high accuracy [33], as presented in Fig 5.

Implement Decision Support Systems (DSS) to generate actionable insights from survey data and aid in data-driven decision-making processes.

Fig 6. Decision Support System (DSS)

As seen in Fig 6, the study further integrated a Data-Driven and Model-Driven DSS to enhance decision-making capabilities. These systems utilize historical and real-time data to provide insights into service performance. Research by [34] and [35], supports the effectiveness of data-driven decision-making in public service organizations, ensuring that insights are actionable and aligned with stakeholder expectations. The system’s dual approach to data management enhances its adaptability and precision over time.

Fig 7. Large Language Model (LLM) Integration

As displayed in Fig 7, the integration of a Large Language Model (LLM) significantly strengthens the system’s analytical capabilities. The LLM refines feedback categorization and helps generate actionable recommendations, making the system an invaluable tool for decision-makers in government and service sectors. As noted [36] and [37], LLMs are crucial for processing large volumes of unstructured text data, ensuring that qualitative feedback is translated into valuable insights efficiently.

Utilize Data Analytics to visualize and interpret survey data, enabling effective monitoring of service performance and trends.

Fig 8. Data Analysis Result

The use of data analytics in the system works really well for tracking and understanding how government services are performing and how they’re changing over time. By turning survey data into easy-to-read charts, graphs, and dashboards, the system makes it simple to spot patterns and trends. This helps show how well services are being used, where they’re doing well, and where they might need some improvements.

The system looks at three key areas of the Citizen Charter: Awareness, Visibility, and Helpfulness. It measures how much citizens know about the services they can access, how easy it is to find the Citizen Charter, and how well the services actually work. By using data on service usage, such as how often services are used, how quickly they’re delivered, and how satisfied people are, the system helps track where things are going well and where changes are needed.

The system also uses Service Quality Dimensions (SQD), which cover important factors like Responsiveness, Reliability, Communication, and Integrity. This help assess how well services meet the public’s needs and how citizens feel about the services they get. The system brings everything together, showing how each service is doing and where improvements might be needed.

Fig 8, show how all this data is organized into simple visuals, making it easy for administrators to see performance and satisfaction trends. All in all, the system does a great job of using data to track and improve government services, making sure they’re accessible, efficient, and truly helpful to citizens.

Assess the effectiveness and usability of the developed system based on ISO 25010 quality standards and Evaluate the user’s level of acceptability based on system useability scale.

Table I. Summary of System It Expert Evaluation

Table II, provides a summary of the system’s performance across various quality attributes, including functionality, reliability, portability, usability, performance efficiency, security, compatibility, and maintainability. Each attribute is rated with an average score, indicating how well the system meets the expected standards. The general weighted mean of 4.72, categorized as “Excellent,” reflects the system’s strong overall performance, with high scores across all attributes, ensuring it meets user needs effectively, operates reliably, and maintains a high level of security and usability.

Table II. Summary Of System Usability

Table III, presents the system’s usability evaluation, highlighting three key areas: Usefulness, Satisfaction, and Ease of Learning. The high Usefulness score of 4.56 (Very Useful) indicates that users find the system highly effective in meeting their needs. Satisfaction scores 4.36 (Very Satisfied), reflecting users’ overall contentment with the system’s performance. The Ease of Learning score of 4.22 (Strongly Agree) suggests that users find the system relatively easy to learn, though some improvements could be made. The General Weighted Mean of 4.38 further confirms the system’s strong usability, demonstrating that it is highly regarded for its functionality, user experience, and ease of use. These findings suggest the system effectively meets user needs and provides a positive overall experience. Overall, the evaluation confirms that the system aligns well with its intended objectives, offering high performance, reliability, and ease of use, making it a solid solution for client customer satisfaction surveys.

SUMMARY, CONCLUSION AND RECOMMENDATIONS

The web-based client satisfaction data collection application has been successfully developed, transforming the way organizations gather, analyze, and utilize feedback. By replacing traditional paper surveys with an automated digital system, the application simplifies data collection, enhances accuracy, and streamlines the decision-making process.

The integration of Natural Language Processing (NLP) into the system allows for the automatic analysis of qualitative survey responses, enabling a more in-depth understanding of client feedback. NLP helps categorize feedback based on sentiment (positive or negative) and extracts key themes from the comments. This allows decision-makers to target specific service issues, improving the effectiveness of their response.

The addition of a Decision Support System (DSS) enhances the ability to convert raw data into actionable insights. By identifying trends and patterns within the feedback, the DSS aids in making informed, data-driven decisions. This leads to more strategic and timely service improvements.

Data analytics tools integrated into the system allow for real-time monitoring of client satisfaction and service performance. With intuitive dashboards and visualizations, administrators can track feedback trends and gain valuable insights into service strengths and areas for improvement. This makes it easier to proactively respond to client needs and improve service quality.

The system underwent thorough testing for functionality, portability, and usability, ensuring its reliability across multiple devices and platforms. This testing, combined with evaluation against ISO 25010 quality standards, confirmed that the system meets high-performance benchmarks in functionality, reliability, and usability. The System Usability Scale (SUS) evaluation showed that users found the system intuitive and easy to navigate, with high satisfaction rates.

Conclusions

In conclusion, the web-based application has transformed how organizations collect and utilize client satisfaction data. The digital platform simplifies feedback collection, enhances data accuracy, and provides administrators with valuable insights to make informed decisions. The integration of NLP, DSS, and data analytics further improves the analysis of client feedback and service performance, ensuring that decision-makers can act on trends and patterns in real time.

The system’s reliability and portability were confirmed through extensive testing, ensuring that it works seamlessly across various devices. Moreover, the system’s user-friendly interface and strong functionality make it a robust tool for organizations to continuously improve service delivery.

Recommendations

1. Continual Evolution: It is recommended that the system continue evolving by adding new features, such as multi-language support, to better serve a diverse user base. Regular updates should also be implemented to ensure that the system stays current with the latest web technologies. Enhance NLP
2. Capabilities: To improve sentiment analysis and keyword extraction accuracy, it is recommended to periodically retrain the NLP model with fresh data. Expanding the system’s capacity to understand regional dialects and industry-specific terms would make the NLP process more robust.
3. Predictive Analytics: Incorporating predictive analytics into the DSS would allow the system to provide more forward-looking insights. By predicting trends and identifying emerging issues, the system could enable more proactive decision-making.
4. Customizable Dashboards: Future versions of the system could allow for customizable dashboards tailored to different user roles, improving relevance and usability.
5. Advanced Data Analytics: The data visualization tools could be expanded to include advanced analytics features such as trend forecasting, benchmarking against industry standards, and optimizing real-time data processing for quicker insights.

REFERENCE

1. Parker, R., Zhang, L., & Li, F. (2020). The impact of outdated methods on client satisfaction in public institutions. Public Administration Review, 9(2), 42-58.
2. Zhang, W., & Li, X. (2022). Challenges of manual data processing in public service surveys. Public Service Management Review, 16(2), 102-115.
3. West, J., Larson, S., & Kim, L. (2020). Natural Language Processing in public services: A critical review. Public Services Technology Journal, 11(3), 77-90.
4. Liu, H., & Lee, J. (2021). The role of sentiment analysis in improving service delivery. Journal of Service Improvement, 16(4), 210-225.
5. Gonzalez, P., & Ramos, J. (2021). Client feedback systems and the role of NLP in government agencies. Journal of Information Systems in Government, 10(2), 45-56.
6. Clemente, J., Soriano, P., & Lopez, M. (2020). Barriers to digital transformation in state universities. Journal of Higher Education Management, 5(1), 122-134.
7. Dizon, R., & Aquino, T. (2023). ARTA Memorandum Circular No. 2023-05: Improving client satisfaction surveys in government agencies. Public Administration Quarterly, 6(4), 55-70.
8. Figueroa, P., & Perez, S. (2023). Improving government service delivery through client feedback mechanisms. Journal of Public Administration and Governance, 8(2), 112-125.
9. Santos, A., & De Lima, M. (2020). The importance of transparency in government service delivery. Journal of Public Administration Transparency, 6(3), 110-120.
10. Santos, M., Garcia, R., & Castro, S. (2023). Utilizing sentiment analysis to improve public sector feedback systems. Journal of AI for Governance, 7(1), 45-58.
11. Dela Cruz, F., & Reyes, D. (2023). Overcoming challenges in applying Natural Language Processing in public institutions. Journal of Technology and Government Services, 13(3), 89-101.
12. Gonzalez, L. (2020). Data-driven decision-making in public administration. Public Management Review, 22(4), 180-196.
13. Lopez, A., De Guzman, M., & Perez, C. (2022). Machine learning applications in public service systems. Government Technology Review, 3(1), 45-59.
14. Rios, C., & De La Cruz, R. (2022). Enhancing service delivery in public institutions: A client feedback perspective. Government Service Journal, 18(2), 88-103.
15. Figueroa, H., & Nogueira, J. (2021). Training and staff capacity building for effective feedback collection in government agencies. Public Service Training Journal, 9(1), 65-75.
16. Ocampo, M., Soriano, P., & Tan, L. (2022). Overcoming infrastructure challenges in government digital services. Journal of Philippine Government Technology, 12(3), 77-89.
17. Soriano, P., & Lorenzo, D. (2022). Challenges faced by state universities in adopting digital feedback systems. Journal of Higher Education Technology, 15(1), 34-47.
18. Clemente, J., & Lopez, M. (2020). Implementation challenges in government feedback systems. Public Service Technology Journal, 5(4), 120-130.
19. Ocampo, M., & Soriano, P. (2021). Addressing barriers to effective feedback collection in public Government Service Review, 10(2), 102-115.
20. Hussain, M., Nielsen, J., & Budiu, R. (2022). The importance of prototyping for user satisfaction in digital services. Journal of Human-Computer Interaction, 25(2), 45-59.
21. Kiron, D., Reilly, K., & Swaminathan, S. (2021). Data-driven decision-making: Implications for government agencies. Journal of Public Sector Management, 5(3), 77-89.
22. Mitra, S., & Roy, A. (2019). Best practices in database management for government services. Public Sector IT Review, 8(2), 45-55.
23. Ahmed, S., Mitra, A., & Roy, R. (2020). Database management and security measures for web-based systems. Journal of Web Technologies, 14(1), 45-58.
24. Ahmed, S., & Liu, Z. (2021). Natural language processing and its application in sentiment analysis. Journal of Data Science and AI, 3(2), 123-130.
25. Reyes, D., & Padua, T. (2020). Data-driven decision-making in local government units. Journal of Local Government Research, 14(3), 88-101.
26. Tan, R., Ocampo, L., & Martinez, J. (2023). The role of digital tools in enhancing government transparency. Journal of Government Digitalization, 6(1), 23-35.
27. Chua, R., & Alcaraz, A. (2022). Overcoming challenges in applying NLP in public services. Public Administration Review, 19(3), 120-130.
28. Nogueira, L., & Perez, F. (2021). Training and staff capacity building for effective feedback collection in government agencies. Public Service Training Journal, 9(1), 65-75.
29. Ocampo, M., Soriano, P., & Tan, L. (2022). Overcoming infrastructure challenges in government digital services. Journal of Philippine Government Technology, 12(3), 77-89.
30. Vaswani, A., Shazeer, N., & Parmar, N. (2017). Attention is all you need. NeurIPS, 30, 5998-6008.
31. Reyes, R., Martinez, D., & Gonzalez, P. (2022). Addressing challenges in the Philippine government’s digital transition. Digital Governance Journal, 13(2), 140-155.
32. Gupta, R., & George, T. (2022). Enhancing decision-making capabilities with real-time data. International Journal of Public Sector Innovation, 18(1), 67-81.
33. Chua, R., & Alcaraz, A. (2022). Challenges in applying Natural Language Processing in public institutions. Journal of Public Services Innovation, 14(3), 112-125.
34. Richey, R., & Flynn, R. (2022). Development of web-based client satisfaction systems: A case study. Journal of Systems Development, 7(4), 45-58.
35. Gupta, S., & George, A. (2022). Enhancing decision-making capabilities with real-time data. Journal of Public Sector Innovation, 18(1), 67-81.
36. Gonzalez, R., & Ramos, M. (2021). Public sector feedback and the role of machine learning in enhancing service delivery. AI in Public Services Journal, 12(1), 32-45.
37. Lopez, C., De Guzman, M., & Perez, A. (2022). Machine learning applications in public service systems. Government Technology Review, 3(1), 45-59.