Arman Process Journal (APJ)

Abstract Background and Objectives: Rapid technological developments in recent decades, particularly in the field of artificial intelligence, have fundamentally transformed traditional leadership models in government organizations. Leaders in the digital age can no longer rely solely on conventional management skills; they must harness the potential of new technologies for intelligent decision-making, creativity, foresight, and effective accountability. While AI has been increasingly adopted in public administration globally, there remains a significant gap in the academic literature regarding how AI specifically contributes to developing leadership skills among government managers, especially in national contexts with unique institutional and cultural characteristics such as Iran. This study aims to explain the role and functions of artificial intelligence in developing leadership skills among managers in Iranian government organizations.
Methods: The present study used a mixed method (qualitative-quantitative) to explain the role and functions of artificial intelligence in developing the leadership skills of managers in Iranian government agencies. In the qualitative stage, by conducting semi-structured interviews with 12 experts in the fields of public administration and information technology, the effective components and indicators were identified.
Findings: Through content analysis, four main dimensions were extracted including digital leadership self-awareness, smart decision-making, smart communication, and AI-based learning and development. In the quantitative phase, to empirically validate the AI-based leadership model developed in the qualitative section, a researcher-made questionnaire was distributed among 80 government managers and the data was analyzed using structural equation modeling.
Conclusion: Analyses show that targeted use of intelligent technologies, especially in the areas of data analysis, organizational behavior prediction, and personalized training, can significantly increase the level of decision-making, managerial agility, and communication capabilities of managers. Also, the lack of data infrastructure and traditional organizational culture are among the most important challenges to the effective use of artificial intelligence in the public sector.

Abstract The present study was conducted with the aim of designing a governance model for big data contracts in the transition to a national digital economy.
This study is applied in terms of purpose and descriptive-exploratory in terms of method, using structural-interpretive modeling (ISM) and MICMAC analysis. The statistical population of the study was composed of experts in the fields of digital economy, information technology law, public policy, and data management, of which 18 out of 35 people were selected through purposive sampling based on theoretical saturation. Data were collected by studying upstream documents, reviewing research literature, and completing paired comparison questionnaires. The content validity of the tool was confirmed by experts, and its reliability was also examined using the intragroup agreement index and the inconsistency rate. Data analysis was conducted using the structural-interpretive modeling approach and MICMAC analysis.
The findings showed that integrated data infrastructure, legal and regulatory frameworks for big data contracts, information security and privacy, transparency and accountability, and institutional coordination are among the driving and fundamental components in the governance of big data contracts. The results of the MICMAC analysis also indicated that compliance with national macro policies and institutional regulation are among the main influential and driving variables of this system.
Accordingly, the establishment of an integrated governance model for big data contracts can pave the way for the realization of a national digital economy by strengthening data-driven decision-making, improving the quality of regulation, increasing transparency, reducing corruption, and improving the efficiency of economic policymaking. Therefore, the proposed research model can be used by policymakers and managers as a strategic framework for designing institutional, legal, and technological mechanisms in the management of data-driven contracts and developing the country’s digital economy.

Abstract This study investigates the impact of social networks on the social security of young people in Semnan Province and proposes a comprehensive framework for explaining the multiple dimensions of this relationship. The research is applied in nature and employs a descriptive–analytical approach. The statistical population consisted of 500 young people aged 15–40 years living in Semnan Province, who were selected through simple random sampling. Data were collected using a researcher-developed questionnaire with confirmed content validity and satisfactory reliability (Cronbach’s alpha > 0.70) and were analyzed using SPSS software.
The results revealed a significant negative relationship between the extent of social network use and different dimensions of social security. Quantitative findings indicated a significant negative correlation between the duration of social network use and academic security (r = −0.42, p < 0.01). Moreover, educational use of social networks had a positive effect on psychological–emotional security (β = 0.27, p < 0.05), whereas entertainment-oriented use showed a significant negative effect (β = −0.31, p < 0.01). Purposeful use of social networks was also associated with higher levels of social participation and psychological well-being, while excessive use was linked to reduced family security, poorer academic performance, and lower social trust.
The main contribution of this study is the development of an integrated framework that simultaneously examines the family, academic, psychological–emotional, and communicative–social dimensions of youth social security while considering the type, intensity, and purpose of social network use. The proposed framework provides a balanced understanding of both the opportunities and the risks associated with social networks and offers practical implications for designing media literacy programs, supporting evidence-based policymaking, and developing preventive strategies aimed at enhancing the social security of young people.

Abstract The rapid proliferation of generative AI models, large language models, coding assistants, AIOps systems, and intelligent cybersecurity tools has transformed ICT architectures and processes. This research aims to provide a technical and analytical examination of the benefits, drawbacks, risks, and applications of prevalent AI tools across the ICT system lifecycle. Adopting a review-analytical approach grounded in authoritative literature, the study first categorizes these tools into domains such as natural language processing, software engineering, data analysis, networking, cloud computing, IT operations, and cybersecurity. It then compares them based on criteria including accuracy, productivity, security, privacy, cost, scalability, integration capabilities, and the need for human oversight. The results indicate that the value of AI in ICT lies less in mere response generation and more in its capacity to analyze vast datasets, automate repetitive tasks, support decision-making, and establish intelligent feedback loops connecting data, software, and infrastructure. Conversely, challenges such as model hallucinations, the generation of vulnerable code, data leakage, prompt injection attacks, vendor lock-in, and decision-making errors constitute significant barriers to widespread adoption. Finally, the study proposes a multi-criteria framework for AI tool selection and a five-level maturity model for organizational deployment. The key finding is that for critical applications, a "human-in-the-loop" architecture, data governance, access control, continuous model evaluation, and an enterprise AI gateway must be integral architectural requirements.

Abstract The expansion of online social networks has attracted a vast number of users. Link prediction is a crucial aspect of social network analysis; its objective is to identify potential connections within a network that are not currently present or observed. In social networks, there is a potential for links to exist between entities—links that may emerge in the future or were simply missing due to data loss. While link prediction is not a form of social network analysis in itself, it is a vital component of the analytical process. Furthermore, recommending interesting friends—in a way that pleasantly surprises new users—has become a major challenge for social networks. This challenge is particularly significant due to both the massive volume of data and the lack of historical activity data for new users upon joining. This paper aims to design a hybrid recommendation system—leveraging novel knowledge and ensemble learning—to facilitate friend recommendations and enhance the accuracy and precision of these suggestions within social networks.

Abstract Sentiment analysis is a key application of natural language processing and artificial intelligence, aimed at identifying and extracting attitudes, emotions, and polarity from human-generated text. Although deep learning models—particularly Transformer-based architectures—have significantly enhanced the capabilities of sentiment analysis systems, analyzing Persian text remains fraught with challenges such as writing variations, the use of zero-width non-joiners (ZWNJ), character form variations, semantic ambiguity, sarcasm, negation, multi-aspect perspectives, a scarcity of labeled data, and a mismatch between model confidence and actual prediction accuracy. Furthermore, many existing systems merely output a general positive, negative, or neutral label without accounting for their confidence in the decision. To address these challenges, this research proposes a multi-stage, context-aware, and uncertainty-driven framework for Persian sentiment analysis. This framework comprises four main layers: Persian preprocessing and normalization, deep representation extraction using Transformer models, aspect-based sentiment analysis, and finally, calibration and uncertainty estimation. This structure enables the distinction between "overall sentiment" and "sentiment regarding a specific aspect." Evaluation results demonstrate that the proposed framework outperforms baseline models across all metrics, including Accuracy, Precision, Recall, and Macro-F1. Additionally, an SCCI value of 91% indicates a high level of consistency and coherence in the model's ability to identify aspect-level sentiment within Persian texts. These results demonstrate that employing the proposed architecture has led to significant improvements not only in classification performance but also in the balanced and simultaneous detection of emotion labels.