Garuda - Garba Rujukan Digital

Article Per Year (5 Year)

All

The Eastasouth Journal of Information System and Computer Science

esiscs
Website

Published by Eastasouth Institute

ISSN : 30266041 EISSN : 3025566X DOI : https://doi.org/10.58812/esiscs

Core Subject : Science,

Computer Science & IT

ESISCS - The Eastasouth Journal of Information System and Computer Science is a peer-reviewed journal and open access three times a year (April, August, December) published by Eastasouth Institute. ESISCS aims to publish articles in the field of Enterprise systems and applications, Database management systems, Decision support systems, Knowledge management systems, E-commerce and e-business systems, Business intelligence and analytics, Information system security and privacy, Human-computer interaction, Algorithms and data structures, Artificial intelligence and machine learning, Computer vision and image processing, Computer networks and communications, Distributed and parallel computing, Software engineering and development, Information retrieval and web mining, Cloud computing and big data. ESISCS accepts manuscripts of both quantitative and qualitative research. ESISCS publishes papers: 1) review papers, 2) basic research papers, and 3) case study papers. ESISCS has been indexed in, Crossref, and others indexing. All submissions should be formatted in accordance with ESISCS template and through Open Journal System (OJS) only.

Arjuna Subject : Ilmu Komputer (semua kategori) - Sistem Informasi

Articles 198 Documents

15 16 17 18 19

GovernedRAG: Risk-Adaptive Retrieval-Augmented Generation for Enterprise Compliance Decision Support Varsha Shah
The Eastasouth Journal of Information System and Computer Science Vol. 2 No. 03 (2025): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v2i03.1229

When enterprises rely on LLMs for compliance decision support because they are under sustained regulatory scrutiny, retrieval-augmented generation without control allows access-sensitive regulatory, contractual, and audit content to be surfaced indiscriminately, and it does not tailor its response to the risk level of the query. This paper combines the latest research in retrieval-augmented generation, access control, adversarial threats, and AI risk management and introduces GovernedRAG, an architecture for enterprise compliance decision support that adapts to risk. The architecture integrates a four-stage risk classifier, a retrieval gate that limits the scope of candidate evidence retrieved to the scope of the requesting user before they are ranked, a grounded generation stage with an evaluation loop based on the reference-free assessment frameworks, and a continuous evaluation loop. A formalization is provided for the risk-weighted relevance, access gating, and faithfulness scoring, and the proposal is compared to five related retrieval-augmented paradigms, and the proposal is juxtaposed with them. It shows that governance and risk adaptivity are still not considered as first-class architectural concerns in previous retrieval-augmented systems, that access control is still mostly outside of retrieval ranking and not embedded in the ranking function, and that reference-free evaluation protocols are practical foundations for ongoing, compliance-driven monitoring. The proposed framework is organized into four risk tiers, four architectural stages and six comparative synthesis tables. Consequences for auditing, regulation and empirical confirmation are considered.

Agentic Enterprise AI Architecture for Autonomous CRM Decision Orchestration Using Salesforce Agentforce and AWS Swapna Putti
The Eastasouth Journal of Information System and Computer Science Vol. 3 No. 01 (2025): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v3i01.1230

Enterprise CRM platforms have evolved from rule-based to autonomous, reasoning agents. In this paper, the authors integrate the insights from the aforementioned twenty publications into an agentic enterprise architecture for autonomous decision orchestration for a CRM system based on Salesforce Agentforce and Amazon Web Services (AWS) large language model (LLM) based multi-agent system. The synthesis looks at an evolution across three generations – deterministic RPA, predictive AI-CRM and agentic AI-CRM – and correlates the generations with the autonomy, latency and human intervention metrics. A layered architecture is proposed with data ingestion layer, predictive personalisation layer, agent orchestration layer, reasoning and planning layer, explainability and governance layer, and finally a cloud infrastructure layer with AWS services being deployed at each layer. Literature review reveals that implementing AI-powered CRM systems is linked to tangible improvements in customer retention, first-response time, and predictive customer service, with agentic architectures also contributing to a decrease in human escalations compared to traditional automation. The major hurdles to enterprise-wide deployment are security, privacy and explainability. The paper introduces a reusable reference architecture, a comparative evaluation framework with confidence-based escalation and autonomy-index formulations, and a set of governance recommendations for organisations that are moving toward using autonomous CRM orchestration. We discuss implications for competitive advantage and operational cost as well as directions for further empirical validation of agentic reasoning loops in production CRM environments.

Reproducibility and Generalization Analysis of OCAT Net S for Oral Cancer Image Classification Tajul Islam Rafi; Md Abu Kawsar Prodhan Hemal
The Eastasouth Journal of Information System and Computer Science Vol. 4 No. 01 (2026): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v4i01.1235

Clinical photographs could support earlier referral of suspicious oral lesions, but high internal accuracy does not show whether a model will remain reliable when image source, disease prevalence, or lesion definition changes. This study conducted a secondary experimental analysis of the aggregate measurements reported for OCAT-Net-S by [1]. The analysis examined internal discrimination, component ablation, synthetic perturbation tolerance, computational efficiency, probability calibration, threshold behavior, fold-level dependence, leakage controls, and external transfer. No new patient images or model-training runs were introduced. All new results in this paper are derived from the published tables. OCAT-Net-S achieved an internal AUC-ROC of 0.956, sensitivity of 0.942, and oral-cancer F1-score of 0.942 on 1,238 clinical oral images. Its AUC advantage over TinyViT-5M, the strongest evaluated baseline, was 0.015. The external AUC fell to 0.830 on SMART-OM and 0.840 on the Oral Images Dataset, producing absolute generalization gaps of 0.126 and 0.116. These gaps were 8.40 and 7.73 times larger than the internal advantage over TinyViT-5M. Structural ablations caused a mean AUC loss of 0.046, compared with 0.0128 for training-related ablations. Under the most severe Gaussian-noise condition, OCAT-Net-S retained 96.3% of its clean AUC, compared with 93.8% for TinyViT-5M and 90.3% for ResNet-50. The model used 3.55 million parameters and achieved 0.269 AUC per million parameters, although MobileNetV3-Large remained faster and required fewer operations. Calibration was favorable internally, with a Brier score of 0.0419 and ECE-15 of 0.0174. At an assumed 3% screening prevalence, however, estimated positive predictive value ranged from 0.213 to 0.420 across reported operating points. The findings support the internal technical value of texture-frequency spatial gating and compact local-global feature learning. They also show that distribution shift and prevalence exerted a larger practical effect than the margin separating OCAT-Net-S from its strongest internal comparator. Patient-linked multicenter evaluation, fixed external thresholds, prospective workflow testing, and quantitative lesion localization remain necessary before clinical use.

The Cost of Late Governance: A Blast-Radius Model for Pre-Promotion vs. Post-Hoc Policy Enforcement in Autonomous Multi-Agent Data Consumption Jagan Ankam
The Eastasouth Journal of Information System and Computer Science Vol. 1 No. 02 (2023): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v1i02.1244

The autopilot multi-agent system is gaining immense popularity as major language models are built with external tools, shared memory and data services to accomplish tasks with minimal human oversight. This architecture transforms the governance issue in that policy breakdown which propagates beyond the initial agent that takes data. The research paper proposes a conceptual Blast-Radius Governance Model to make a comparison between pre-promotion policy enforcement and post-hoc enforcement in autonomous multi-agent data consumption. It claims that the timing of governance is a risk variable as opposed to administration choice. In this research the suggested model determines the blast radius by considering the affected-agent scope, data exposure, depth propagation, action impact, governance latency and recoverability. The analysis indicates that pre-promotion controls might maintain some limits on unsafe states prior to their reachability, but post-hoc controls do so once they have been exposed to some degree. An integrated pre-promotion validation, runtime monitoring and post-hoc monitoring is proposed as a layered model. The research paper presents a conceptual foundation which can be tested but does not purport to be empirically validated.

From Custom Code Inventory to Clean Core: CORE-Fit—A Portfolio-Level Methodology for Rationalizing ERP Extensions in SAP S/4HANA Abhinav Jain
The Eastasouth Journal of Information System and Computer Science Vol. 3 No. 01 (2025): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v3i01.1245

Enterprise resource planning (ERP) modernization increasingly requires organizations to reconcile two competing objectives: preserving business-specific capabilities and reducing the lifecycle burden created by tightly coupled custom code. Clean Core addresses this tension through fit-to-standard processes, governed extensibility, stable interfaces, and deliberate placement of capabilities inside or outside the transactional ERP core. This paper introduces CORE-Fit (Customization Optimization, Rationalization, and Extensibility Fit), a design-science methodology proposed in this study for rationalizing inherited ERP customizations into five disposition paths: retire or standardize, key-user extension, on-stack developer extension, side-by-side extension, or governed exception and remediation. The underlying principles and related SAP approaches to custom-code rationalization and extension design were already established; CORE-Fit does not claim novelty for retirement, fit-to-standard, or extension technology selection as individual concepts. Its contribution is the specific six-stage portfolio-level decision artifact that integrates business-need validation, standard-fit analysis, transaction and data proximity, lifecycle independence, integration and economic/operational fit, and governance into five explicit disposition outcomes. CORE-Fit is evaluated using an anonymized customer-system assessment of 121 RICEF objects. The resulting disposition comprised 22 objects (18.2%) retired or standardized, 11 (9.1%) classified as key-user extensions, 73 (60.3%) as on-stack developer extensions, 15 (12.4%) as side-by-side extensions, and none as governed exceptions. The results illustrate that Clean Core rationalization need not equate to aggressive externalization: in this portfolio, most surviving requirements remained close to the transactional core through governed developer extensibility. The study presents a case-evaluated methodology for applying these criteria consistently across a legacy customization portfolio.

Cross-Protocol Data-Integrity Validation During Live Migration Between NVMe and Fibre Channel Storage Paths Gagan Vyas
The Eastasouth Journal of Information System and Computer Science Vol. 3 No. 02 (2025): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v3i02.1246

Any data corruption that happens between the two storage paths while they are both active can go unnoticed. The 20 records supplied were used to conduct a systematic review of how data-integrity validation can be organized as the volumes move between Fibre Channel (FC) and Non-Volatile Memory Express (NVMe) paths. Evidence of the effects of storage errors, flash media, fabric performance, protocol digests, and migration mechanics was aggregated and treatment was added to an analytical model of validation cost and residual risk. Field data revealed 0.86% disks with checksum mismatches on nearline disks and 0.065% FC enterprise disks with checksum mismatches; 3.45% disks with latent sector errors. The application throughput was 20% lower with remote flash access over the Internet Small Computer System Interface (iSCSI) than without it, while NVMe over Fabrics introduced 11.7 µs of latency. With a corruption probability of 10⁻⁷ per block, the model predicted that about 6.25×10⁻⁹ blocks would be corrupted each time 1 TiB was migrated, but, as the 4 GiB/s full inline validation showed, pre-copy migration was not convergent. A compromise was reached by sampling the digest on the fly during the copying process and then comparing the final digest with the on disk digest. The main research gap identified is having no direct empirical study on FC-to-NVMe migration.

Architecting Responsible AI Governance on SAP BTP for Intelligent Enterprise Transformation through Explainable Systems, Adaptive Compliance Controls, and Trusted Automation Engineering Rajendra Asuri
The Eastasouth Journal of Information System and Computer Science Vol. 3 No. 02 (2025): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v3i02.1247

Rapid enterprise adoption of artificial intelligence within SAP Business Technology Platform environments has introduced critical challenges associated with governance transparency, ethical automation, regulatory compliance, and operational accountability. Existing enterprise governance models often lack integrated mechanisms capable of addressing explainability, adaptive policy enforcement, and intelligent risk monitoring across complex AI driven ecosystems. This study proposes a comprehensive Responsible AI Governance framework for SAP BTP deployments designed to strengthen trusted enterprise transformation through explainable systems, adaptive compliance controls, and secure automation engineering. The research applies a mixed methodological approach that combines qualitative analysis of enterprise governance practices with quantitative evaluation of governance effectiveness across AI lifecycle management processes. The framework integrates explainable intelligence models, automated compliance orchestration, policy driven governance workflows, and continuous risk assessment mechanisms to improve transparency, operational resilience, and decision accountability within enterprise platforms. Findings demonstrate that the proposed architecture significantly enhances governance efficiency, reduces compliance complexity, improves trust in AI enabled business operations, and supports scalable enterprise innovation without compromising security or ethical standards. The study contributes strategically by establishing a governance blueprint that aligns enterprise AI adoption with organizational accountability and regulatory readiness, while academically extending existing research in responsible AI governance and intelligent enterprise systems. The proposed framework provides a scalable foundation for organizations seeking to operationalize trustworthy AI capabilities within modern SAP driven digital transformation environments.

A Privacy-Preserving Federated Learning Framework on SAP Business Technology Platform for Secure, Scalable, and Collaborative Predictive Analytics in Distributed Enterprise Environments Rajendra Asuri
The Eastasouth Journal of Information System and Computer Science Vol. 2 No. 01 (2024): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v2i01.1248

Enterprise organizations increasingly rely on predictive analytics while facing strict data privacy requirements and fragmented data landscapes that limit centralized model training. This study addresses the challenge of enabling accurate and scalable predictive intelligence without exposing sensitive enterprise data by proposing a privacy-preserving federated learning framework implemented on SAP Business Technology Platform. The research problem focuses on balancing model performance, data security, and interoperability across distributed environments. A mixed-method approach is adopted, combining quantitative evaluation of model accuracy, convergence behavior, and communication efficiency with qualitative assessment of governance, trust, and deployment feasibility. The proposed framework integrates secure aggregation, decentralized training orchestration, and policy-driven data access controls to ensure privacy while maintaining analytical performance. Experimental results demonstrate that federated models achieve comparable accuracy to centralized approaches while significantly reducing data exposure risks and improving compliance alignment. The study introduces an architectural blueprint that enhances scalability, resilience, and cross-organizational collaboration. Findings contribute to both academic discourse on distributed machine learning and industry practices by establishing a structured approach to privacy-aware predictive analytics. The research concludes that federated learning on SAP BTP enables sustainable, secure, and enterprise-ready AI adoption, offering strategic value for organizations seeking to unlock insights from sensitive and distributed data ecosystems.

Deadline Performance and Tenant Fairness Under Security Screening in Cloud Workload Scheduling Kesavan Sundara Mudaliyar; S. Satish Kumar
The Eastasouth Journal of Information System and Computer Science Vol. 3 No. 01 (2025): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v3i01.1250

Cloud applications share execution capacity while imposing different deadlines and security screening requirements. A scheduler can therefore face congestion even when aggregate capacity appears sufficient, particularly when a large tenant reaches its concurrency limit. This study investigates interaction through a reproducible discrete event simulation of screened cloud jobs. Four nonpreemptive policies are compared: first in first out, shortest job first, earliest deadline first, and a deadline policy with a bounded penalty for tenants that have already consumed more virtual service. The primary experiment contains 240 runs covering three arrival rates, 20 random seeds, and 1,200 jobs per run. An additional 80 runs remove the restrictive tenant concurrency limit at the highest arrival rate. With eight workers and a four-worker tenant limit, shortest job first has a mean deadline miss rate of 5.26% at 1.5 jobs per simulated second, compared with 44.57% for first in first out and 23.97% for earliest deadline first. The service balance candidate reaches 28.91%, showing that the added fairness penalty does not improve deadline performance in this workload. Increasing the tenant limit from four to eight reduces the earliest deadline miss rate to 1.41%. These findings identify tenant capacity fragmentation and job length variation as stronger influences than the tested fairness adjustment. The results are generated from synthetic workloads and do not measure production cloud performance or resistance to cyberattacks. The contribution is an executable comparative study of scheduling under a mandatory modeled screening stage, with explicit workload assumptions and a tenant capacity sensitivity analysis.

AI Governance: Automatically Checking AI Systems Against the EU AI Act and NIST Rules Gopinathan Rathinavelu
The Eastasouth Journal of Information System and Computer Science Vol. 2 No. 03 (2025): The Eastasouth Journal of Information System and Computer Science (ESISCS)
Publisher : Eastasouth Institute

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58812/esiscs.v2i03.1257

Financial institutions deploying models for credit evaluation, fraud detection, and anti-money laundering must demonstrate that their internal documentation satisfies frameworks such as the European Union's Artificial Intelligence Act (EU AI Act) and the NIST AI Risk Management Framework (AI RMF). Automating this documentation review is attractive, but a checker that approves a control merely because a document mentions the right topic keywords is dangerous: it converts empty corporate assurance into apparent compliance. We present a system design and evaluation for evidence-linked compliance checking, with three contributions. First, a structured control catalog of twelve documentation-checkable controls mapped to EU AI Act articles and NIST AI RMF functions, each with explicit substance criteria. Second, an evidence-linked checker architecture whose every "satisfied" decision is tied to an exact supporting sentence, supported by a tiered decision pipeline, calibrated abstention, and a reviewer workflow. Third, a controlled synthetic benchmark of financial AI documentation in which each control is substantive, boilerplate-only, or absent, with disjoint training and unseen phrasing sets. In our baseline experiments, keyword rules falsely approve 85% of boilerplate-only controls. Trained sparse lexical classifiers reach high precision but, under unseen phrasing, recall falls to between 0.12 and 0.22. Adding 25 labeled documents in the new phrasing restores recall above 0.99, an estimate we caution is optimistic given the synthetic setting. Removal-based deletion tests show that evidence-faithfulness metrics must be interpreted against the redundancy of evidence within a document. The benchmark is a measurement tool for checker reliability, not a legal determination of conformity.

Page 17 of 20 | Total Record : 198

15 16 17 18 19

Filter by Year

2023 2026

Filter By Issues

Home Page

OAI Link

Editorial Team

Contact

Reviewer

Google Scholar

Contact Name
Abdul Aziz

Contact Email
abdulazizbinceceng@gmail.com

Phone
+6282180992100

Journal Mail Official
journaleastasouth@gmail.com

Editorial Address
Grand Slipi Tower, level 42 Unit G-H Jl. S Parman Kav 22-24, RT. 01 RW. 04 Kel. Palmerah Kec. Palmerah Jakarta Barat 11480

Location
Kota adm. jakarta barat,

Dki jakarta

INDONESIA

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Home Page

OAI Link

Editorial Team

Contact

Reviewer

Google Scholar

Contact Name Abdul Aziz

Contact Email abdulazizbinceceng@gmail.com

Phone +6282180992100

Journal Mail Official journaleastasouth@gmail.com

Editorial Address Grand Slipi Tower, level 42 Unit G-H Jl. S Parman Kav 22-24, RT. 01 RW. 04 Kel. Palmerah Kec. Palmerah Jakarta Barat 11480

Location Kota adm. jakarta barat, Dki jakarta INDONESIA

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Contact Name
Abdul Aziz

Contact Email
abdulazizbinceceng@gmail.com

Phone
+6282180992100

Journal Mail Official
journaleastasouth@gmail.com

Editorial Address
Grand Slipi Tower, level 42 Unit G-H Jl. S Parman Kav 22-24, RT. 01 RW. 04 Kel. Palmerah Kec. Palmerah Jakarta Barat 11480

Location
Kota adm. jakarta barat,

Dki jakarta

INDONESIA