نوع مقاله : مقاله پژوهشی

نویسنده

• جواد مقتدر کارگران

دکتری مدیریت، دانشکده اقتصاد و مدیریت، دانشگاه آزاد اسلامی تبریز، تبریز، ایران

چکیده

در عصر تحول دیجیتال، مدیریت دانش به‌عنوان یک منبع کلیدی سازمانی نقشی اساسی در ارتقای تصمیم‌گیری، چابکی سازمانی و توسعه سرمایه فکری ایفا می‌کند. هدف این پژوهش طراحی چارچوبی بومی برای به‌کارگیری هوش مصنوعی مولد در مدیریت دانش سازمان‌های دولتی ایران است. این پژوهش از نوع کاربردی–توسعه‌ای بوده و با رویکرد کیفی و استفاده از روش نظریه‌پردازی داده‌بنیاد انجام شده است. داده‌ها از طریق مصاحبه‌های نیمه‌ساخت‌یافته با ۱۵ نفر از خبرگان حوزه‌های مدیریت دانش، فناوری اطلاعات و تحول دیجیتال در دستگاه‌های دولتی گردآوری و با نرم‌افزار مکس‌کیودا در سه مرحله کدگذاری باز، محوری و انتخابی تحلیل شد. یافته‌ها به شناسایی پنج مقوله اصلی شامل زیرساخت‌های فناورانه، مدیریت کیفیت دانش مولد، فرهنگ‌سازمانی یادگیرنده، حاکمیت داده و اخلاق هوش مصنوعی، و نظام بازخورد و بهبود مستمر منجر شد که درمجموع چارچوب مفهومی پیشنهادی برای استقرار مدیریت دانش مبتنی بر هوش مصنوعی مولد در سازمان‌های دولتی را شکل می‌دهند. این چارچوب می‌تواند به سیاست‌گذاران و مدیران بخش دولتی در بازطراحی نظام‌های مدیریت دانش با بهره‌گیری از قابلیت‌های هوش مصنوعی مولد کمک کرده و به بهبود کیفیت تصمیم‌گیری، افزایش چابکی سازمانی و ارتقای شفافیت منجر شود.

کلیدواژه‌ها

• مدیریت دانش
• هوش مصنوعی مولد
• سازمان‌های دولتی
• نظریه داده‌بنیاد
• مدل‌های زبانی بزرگ

موضوعات

• نظام‌ها و فناوری‌های مدیریت دانش

عنوان مقاله [English]

Generative Artificial Intelligence-Based Knowledge Management Framework in Government Organizations

نویسنده [English]

• Javad Moghtader Kargaran

Ph.D. in Management, Faculty of Economics and Management, Islamic Azad University of Tabriz, Tabriz, Iran

چکیده [English]

Introduction

In recent decades, knowledge management has played a key role as one of the strategic pillars of organizations, particularly in the public sector, contributing to performance improvement, effective decision-making, social accountability, and organizational learning )Dalkir, 2017; Alavi & Leidner, 2001(. As organizational environments become increasingly complex and innovation in public service delivery becomes imperative, public organizations must distance themselves from traditional models and move toward intelligent knowledge management systems (Choi et al., 2020).
Amid these developments, one of the most significant recent technological advancements has been the emergence of Generative Artificial Intelligence and Large Language Models such as ChatGPT, Claude, and Gemini, which are capable of generating textual content, performing conceptual inference, and analyzing natural language )Bommasani et al., 2022). These technologies, with capabilities such as information summarization, tacit knowledge extraction, and intelligent responsiveness, hold significant potential for transforming the knowledge management cycle )Koskella et al., 2023).
Therefore, the present research has been conducted with the aim of designing an indigenous and practical conceptual framework for implementing generative AI-based knowledge management in Iranian public sector organizations. Utilizing field data and a grounded theory approach, this study seeks to provide an executable solution for addressing technological, human, and governance challenges in the digital transformation of knowledge management.

Literature Review

Numerous studies have demonstrated that artificial intelligence, particularly generative AI, can enhance knowledge management processes. Chen et al. )2016) established a positive relationship between AI application and knowledge management effectiveness as well as organizational performance. Sivaraja et al. )2017) recognized intelligent technologies as effective in facilitating knowledge management transformation within e-government. Jarrahi et al. )2018) introduced AI as a complement to human decision-making and influential in organizational knowledge creation.
Abubakar et al. (2020) highlighted the combined capabilities of AI and big data in improving organizational knowledge creation, while Chatterjee et al. (2021) emphasized the importance of coordination among technology, policymakers, and users in AI implementation. Rahman et al. )2022( pointed to infrastructural and policy challenges in leveraging AI within the public sector. Additionally, Dwivedi & Koskela )2023( confirmed the capability of large language models such as ChatGPT in accelerating and facilitating knowledge management.
In Iran, studies such as Rahim & Asgari )2020( and Akbari et al. )2022( have addressed cultural, structural, and technological challenges in implementing knowledge management within governmental bodies; however, less attention has been given to the role of generative AI.

Methodology

This study employs a qualitative research approach using grounded theory methodology to extract and develop a conceptual framework for generative AI-based knowledge management in Iranian public-sector organizations. Grounded theory is a systematic and theory-oriented method that generates theory directly from field data, proving particularly effective when theoretical knowledge is limited or ambiguous )Strauss & Corbin, 1998(.
The study population consists of experts, senior managers, and specialists in knowledge management, information technology, and digital transformation within Iranian public sector organizations, all possessing relevant experience and expertise related to the research topic. Theoretical purposive sampling was employed, meaning participants were selectively chosen based on criteria such as specialized knowledge, relevant work experience, and ability to provide rich data pertinent to the research subject. The sampling process continued until theoretical saturation was achieved, with data collection ceasing when no new information or emerging concepts could be extracted. The final sample comprised 15 experts with relevant academic and professional backgrounds, who participated in semi-structured interviews.
Data were collected through semi-structured in-depth interviews. The interview questions were designed to extract detailed perspectives regarding experiences, challenges, opportunities, and expectations related to the application of generative AI in knowledge management. Following informed consent, all interviews were audio-recorded and subsequently transcribed verbatim. Each interview lasted between 45 to 60 minutes.
Data analysis followed the classical grounded theory procedures:
Open Coding: Raw data were examined line-by-line to extract initial concepts. This phase involved breaking down data into constituent elements and identifying fundamental concepts.
Axial Coding: Extracted concepts were categorized and interconnected to form conceptual structures and identify main categories.
Selective Coding: Main categories were integrated to develop the final theoretical framework based on causal and structural relationships between concepts.
The specialized software MAXQDA was utilized for data organization, coding, and analysis, enabling comprehensive management of qualitative data and systematic code classification.

Results

Data analysis revealed that implementing generative AI-based knowledge management in public sector organizations necessitates the presence of advanced technological infrastructure, a learning-oriented culture, ethical and data governance systems, as well as continuous evaluation and feedback mechanisms. The integration of these elements can enhance the effectiveness of knowledge management, improve organizational decision-making, and promote functional transparency within public organizations.

Discussion

Information technology infrastructure serves as the foundational platform for implementing generative AI, playing a vital role in organizational empowerment. Knowledge quality management, with its focus on continuous validation and refinement, ensures the accuracy and applicability of generated content. A learning organizational culture and technological acceptance, coupled with leadership support and ongoing employee training, constitute essential prerequisites for the success of this process. Data governance principles, transparency, and professional ethics enhance public trust and elevate information security. Finally, the establishment of feedback mechanisms and continuous learning enables the framework to dynamically adapt to environmental changes.
These findings align with previous studies in the domains of knowledge management and AI applications, demonstrating the importance of comprehensive and localized approaches in leveraging emerging technologies within the public sector.

Conclusion

The proposed framework from this research can guide governmental organizations in strategically leveraging generative AI capabilities for knowledge management, ultimately enhancing decision-making, organizational agility, and productivity. With its high localizability, this framework enables the development of strategic plans and executive policies for knowledge-driven digital transformation in Iran's public sector. Furthermore, the embedded continuous feedback mechanism facilitates ongoing evaluation and gradual improvement of knowledge management systems.
The findings of this research yield two key categories of implications:
Theoretical Implications:
This study expands the knowledge management literature by incorporating the "generative AI" dimension, demonstrating how large language models can serve as a resource for creating, refining, and organizing organizational knowledge. The proposed framework bridges abstract knowledge management concepts with operational capabilities of generative technologies, paving the way for developing new conceptual models in the future.
Practical Implications:
The presented framework can serve as an implementation guide for managers and policymakers in public sector organizations. It not only facilitates the redesign of governmental knowledge management systems with an emphasis on transparency, agility, and data-driven decision-making but also provides tools for data quality management, enhancing learning organizational cultures, and ensuring ethical considerations in AI deployment. Moreover, the framework's integrated feedback and continuous improvement mechanisms create conditions for organizations to remain flexible and sustainable alongside future technological advancements. Thus, this research offers concrete solutions for policy-making and management in Iran's public sector, in addition to its scientific and theoretical value.

کلیدواژه‌ها [English]

• knowledge management
• generative artificial intelligence
• government organizations
• data-driven theory
• large language models
• Digital transformation