Current Issue

All articles published in this issue have undergone a thorough peer review process, and stringent checks for repetition rates have been implemented to ensure the integrity of the content.

Total number of articles in this issue: 5
Total number of pages in this issue: 38

For inquiries regarding the content of specific articles, please feel free to contact the respective authors via their provided email addresses. For questions related to the journal itself, please reach out directly to SUAS Press.

Year

2026

Volume

1

Number

2

Status

Archived

Published

2026 April 10

Articles

Open Access Research ArticlePublished 10 April 2026

Practice and Challenge of Generative Artificial Intelligence in Architectural Creative Design

Lin Yang

Under the background of the rapid development of science and technology, the relationship between generative AI (artificial intelligence) and the field of architectural creative design has become an important topic that has attracted much attention. With the continuous progress of science and technology, generative AI is gradually infiltrating into architectural design. In-depth study on the application of generative AI in this field is of great significance to promote the development of architectural creative design. In this paper, the theoretical basis of generative AI in the field of architectural creative design is deeply analyzed, and the internal logic of its combination with architectural design is explained in detail. At the same time, this paper focuses on the practical mode of generative AI in building concept generation, spatial layout design and shape shaping. Although generative AI has brought new opportunities for architectural creative design, it also faces many challenges in the process of practical application. Based on this, this paper puts forward a series of targeted strategies.

Disciplines:Intelligent SystemsSubjects:OtherReferences:22

DOI10.70393/6a696574.343130ARKark:/40704/JIET.v1n2a01
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 10 April 2026

Research on Stress Testing Automation of AI Server for High Concurrency Scenarios

Xingcheng Ren

Traditional stress testing methods are difficult to simulate the complexity and dynamics in real business scenarios, resulting in distorted test results and low efficiency. In order to solve the above problems, this paper proposes an automated framework for stress testing of AI servers facing high concurrency scenarios. The framework adopts the design concept of hierarchical decoupling and intelligent decision-making, and consists of four modules: intelligent load generation layer, system resources and performance monitoring layer, dynamic tuning and control center, root cause analysis and report generation layer. Among them, the intelligent load generation layer supports mixed simulation of multi-modal AI loads, the dynamic tuning and control center realizes dynamic optimization of test parameters based on reinforcement learning (RL) algorithm, and the root cause analysis and report generation layer automatically locates performance bottlenecks and generates reports by unsupervised learning and time series correlation analysis. The experimental results show that the framework can effectively find the potential bottlenecks of the system, improve the test efficiency, and shorten the fault diagnosis cycle, which provides strong support for the performance optimization of AI server.

Disciplines:Intelligent SystemsSubjects:OtherReferences:22

DOI10.70393/6a696574.343131ARKark:/40704/JIET.v1n2a02
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 10 April 2026

Collocation Generation and Style Consistency Verification of Personalized Soft Outfit Based on Diffusion Model

Lijun Zhang

There are some problems in traditional soft outfit design, such as low efficiency, difficult style understanding and low user participation, and the development of generative AI has brought new opportunities to solve these problems. In this study, a collaborative soft outfit generation system between users and AI is constructed. Through multi-modal condition fusion and dynamic style constraint method, users' fuzzy preferences are transformed into quantitative conditions, and personalized soft outfit schemes conforming to specific styles are generated. The system uses CLIP text and image encoder to extract features, and controls them in the potential space based on Stable Diffusion architecture, and optimizes the generation effect by combining style contrast loss. At the same time, a three-level verification system is proposed, including quantitative evaluation indicators such as ArtFID and FCS (Furnishing CLIP Score), and human evaluation as a supplement to objectively quantify the matching degree between the generated results and the target style. The experiment uses the FurniSynth data set built by ourselves, covering 10 mainstream decoration styles. The results show that compared with the existing methods, the method proposed in this paper has the best performance in style consistency, with an ArtFID index score of 19.3, a FCS semantic matching degree of 0.83, and a designer's subjective score of 4.6. In addition, the ablation experiment shows that the overall performance of the model can be gradually improved by adding each component in turn. This study not only technically breaks through the control problem of traditional model on style consistency, but also reduces the design cost of small and medium-sized enterprises in industry and improves efficiency, and promotes the integration of personalized consumption mode and smart home aesthetic design in society.

Disciplines:Intelligent SystemsSubjects:OtherReferences:25

DOI10.70393/6a696574.343132ARKark:/40704/JIET.v1n2a03
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 18 May 2026

Research on the Long-Term Mechanism of Digital Transformation for Small and Medium-Sized Enterprises

Nini Zhang

The digital transformation of small and medium-sized enterprises (SMEs) remains perpetually haunted by a paradoxical state where high initial capital outlays often yield negligible short-term returns, leading to a precarious "transformation interruption." Moving beyond simplistic linear assumptions that dominate much of the current literature, this research synthesizes TOE theory, resource-based views, and dynamic capabilities to construct a "four-helix" long-term mechanism, comprising Driving Force, Conduction, Guarantee, and Iteration, to scrutinize the systemic sustainability of digital value creation. Our empirical journey, which necessitated a transition from traditional panel regressions to a more nuanced mixed-method design including Bootstrap chain mediation and fsQCA to account for the messy realities of organizational flux, examines 1,743 A-share listed SMEs from 2018 to 2023. Findings tentatively reveal a significant U-shaped trajectory between digital intensity and performance, suggesting a theoretical "inflection point" at approximately 24–30 months; however, this duration possibly fluctuates based on the maturity of the digital ecosystem. While dynamic capability and business model innovation emerge as vital chain mediators, arguably facilitating the core value transformation. Their effectiveness is perhaps contingent upon a recursive "Iteration" helix that demands further longitudinal validation. Notably, configuration analysis indicates that four-helix synergy achieves a 68.7% success rate, a stark contrast to single-factor drivers, emphasizing that long-term efficacy is a product of configurational alignment rather than isolated technological adoption. These insights lead us to further thinking about the institutionalization of digital habits, suggesting that future SME resilience relies on a self-driven, dynamically iterative system rather than transient software acquisition.

Disciplines:Intelligent SystemsSubjects:Adaptive ControlReferences:35

DOI10.70393/6a696574.343133ARKark:/40704/JIET.v1n2a04
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 18 May 2026

A Comparative Empirical Evaluation of Single-Agent and Multi-Agent LLM Prompting Strategies for Automated Formative Feedback in Education

Jiawen Lai, Zan Li

Automated formative feedback has emerged as a focal point in educational technology research, as large language models (LLMs) offer the prospect of providing personalized commentary on student writing at a scale that human instructors alone cannot match. What is less well examined, however, is how the underlying prompting design—particularly the choice between single-agent and multi-agent setups—shapes the pedagogical value of the feedback produced. To examine this question, we conducted a controlled comparison across four prompting configurations on a corpus of 200 undergraduate argumentative essays: a zero-shot single-agent baseline, a chain-of-thought single-agent variant, a dual-role multi-agent pipeline in which one model drafts feedback and another critiques it, and a tri-role multi-agent pipeline that introduces a dedicated revision stage on top of the draft-and-critique loop. Each set of feedback outputs was assessed along a multi-dimensional rubric covering accuracy, specificity, constructiveness, and tone, with three trained raters scoring independently. We also computed automated textual similarity metrics against expert-authored reference feedback to complement the human ratings and provide a more independent check. The tri-role multi-agent configuration produced the highest composite quality scores and, notably, the lowest rates of over-praise and hallucinated claims about essay content. The chain-of-thought single-agent variant, while not topping the rankings, delivered surprisingly close quality at a fraction of the inference cost, making it an attractive option when computational budget or latency matters. We close by discussing what these patterns mean in practice for educators and developers looking to integrate LLM-based feedback agents into higher-education writing workflows at scale.

Disciplines:Intelligent SystemsSubjects:OtherReferences:28

DOI10.70393/6a696574.343135ARKark:/40704/JIET.v1n2a05
View ArticleDownload PDFScholar

Open Access Research ArticlePublished 22 June 2026

Design of an AI Agent-Driven Long-Tail Intent Perception Framework for Intelligent Customer Service in Low-Resource Environments

Yanyan Zhang

Intelligent customer service platforms have gained widespread deployment in numerous sectors in recent years. Even so, most mainstream systems run into bottlenecks when tackling rare, long-tail user questions. Issues like insufficient labeled data, varied expression habits and inadequate domain knowledge make such problems worse, especially for businesses lacking ample training datasets. Against this practical dilemma, this study puts forward the LTIP framework tailored for long-tail intent identification, which combines agent technology, retrieval-augmented generation, confidence scoring mechanisms as well as adaptive human-agent transfer logic. The framework employs multi-stage intent perception to identify long-tail inquiries, retrieves enterprise knowledge through RAG, and evaluates response reliability using a confidence-aware mechanism. Based on the confidence score, customer requests are dynamically routed to either AI services or human agents. The proposed framework improves long-tail issue handling while reducing dependence on large-scale labeled datasetsThis design delivers a streamlined, workable option to roll out smart customer service tools under constrained data and resource conditions.

Disciplines:Intelligent SystemsSubjects:Autonomous AgentsReferences:28

DOI10.70393/6a696574.343139ARKark:/40704/JIET.v1n2a06
View ArticleDownload PDFScholar

Copyright © 2026 The Author(s). Published by Southern United Academy of Sciences.
This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Journal of Intelligence and Engineering Technology (JIET)
Published by Southern United Academy of Sciences Limited ISNI: 0000000512776460, operated by the Publications Division.