International Journal of Computer Science and Technology

International Journal of Computer Science and Technology | Vol. 4, No. 4, April 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000219

Article Type: Original Research Paper

Title: Enhancing Cross-Lingual Named Entity Recognition via Contextual Word Embedding Alignment

Names of Authors: Priya Nair¹, Hans-Jürgen Integrated²

Authors’ Affiliations: ¹Center for Computational Linguistics, Indian Institute of Technology, Chennai, India; ²Computational Linguistics Division, Munich Language Labs, Munich, Germany

Abstract: Named Entity Recognition (NER) models achieve high accuracy when trained on large, labeled English datasets but perform poorly in low-resource languages. Manually labeling text corpora across multiple languages is expensive and time-consuming. This paper introduces a cross-lingual NER framework that transfers entity knowledge from English to resource-scarce target languages without requiring manual translations. The framework uses a mathematical alignment layer that projects localized target-language text features into a shared semantic vector space alongside pre-trained English models. We incorporated a token-level attention module to handle variations in word order and syntax across languages. The model was validated on Spanish, Hindi, and Turkish text datasets. The experimental results show a mean F1-score improvement of 11.2% over existing unsupervised cross-lingual models. The framework proved highly effective at identifying complex entities like corporate names and geographic locations across different language syntax models. This alignment strategy provides an efficient method to scale multilingual text analytics tools.

Keywords: Named Entity Recognition, Cross-Lingual Transfer, Natural Language Processing, Word Embeddings, Low-Resource Languages, Semantic Alignment

Manuscript Timeline: Received: June 12, 2022; Revised: July 18, 2022; Accepted: August 08, 2022; Published: April 09, 2023

International Journal of Computer Science and Technology | Vol. 4, No. 3, March 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000218

Article Type: Original Research Paper

Title: Optimizing Energy Usage in Smart Grids via Distributed Multi-Agent Reinforcement Learning

Names of Authors: Aleksei Ivanov¹, Chidi Okoro²

Authors’ Affiliations: ¹Power Systems Computing Lab, Saint Petersburg Tech University, Saint Petersburg, Russia; ²Department of Electrical Engineering, University of Ibadan, Ibadan, Nigeria

Abstract: Modern electrical grids struggle to balance dynamic consumer power demands with highly volatile renewable energy sources like wind and solar. Centralized grid management models face communication delays and scalability bottlenecks when coordinating millions of independent endpoints. This paper presents a decentralized energy management framework that uses a Multi-Agent Deep Deterministic Policy Gradient (MADDPG) model. In this setup, independent software agents represent localized residential neighborhoods and regional solar generation grids. These agents learn to coordinate energy storage allocations and balance localized loads by exchanging minimal peer-to-peer state messages. We simulated this multi-agent architecture using a 120-node electrical distribution setup. The empirical results show a 24.3% reduction in peak-hour grid stress along with an 18.5% optimization in local battery utilization. The decentralized architecture maintained operational stability even during sudden 40.0% generation drops caused by shifting weather conditions. This multi-agent coordination approach provides a resilient method to scale green energy integration across municipal utility networks.

Keywords: Smart Grids, Multi-Agent Systems, Reinforcement Learning, Renewable Energy, Load Balancing, Energy Storage

Manuscript Timeline: Received: June 10, 2022; Revised: July 15, 2022; Accepted: August 05, 2022; Published: March 05, 2023

International Journal of Computer Science and Technology | Vol. 4, No. 2, February 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000217

Article Type: Review Paper

Title: Homomorphic Encryption Pipelines for Secure Cloud Analytics: A Performance Review

Names of Authors: Oliver Smith¹, Sophia Rossi²

Authors’ Affiliations: ¹Department of Cryptographic Research, Manchester Advanced Science University, Manchester, UK; ²Data Protection Systems Division, CyberGuard Labs, Milan, Italy

Abstract: Fully Homomorphic Encryption (FHE) offers a robust privacy framework by enabling mathematical operations directly on encrypted data payloads. This capability allows cloud servers to process sensitive data fields without decrypting the underlying source records. This review paper analyzes the evolution of FHE implementation frameworks over the past decade. We contrast lattice-based mathematical structures including BGV, BFV, and CKKS schemes across cloud environments. The analysis tracks key performance metrics including key sizing requirements, processing overhead factors, and noise growth rates. We compiled empirical benchmarks from 32 peer-reviewed implementations across standard business operations. The evaluation indicates that while CKKS optimizes floating-point operations for machine learning workloads, processing demands remain up to 100x slower than unencrypted workflows. The review outlines hardware acceleration strategies using modern FPGA and ASIC chipsets to mitigate these performance bottlenecks. Finally, we provide a structured configuration guide to help developers choose FHE frameworks based on specific operational requirements.

Keywords: Fully Homomorphic Encryption, Cloud Computing, Data Privacy, Lattice-Based Cryptography, Performance Benchmarks, Hardware Acceleration

Manuscript Timeline: Received: June 05, 2022; Revised: July 12, 2022; Accepted: August 03, 2022; Published: February 11, 2023

International Journal of Computer Science and Technology | Vol. 4, No. 1, January 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000216

Article Type: Original Research Paper

Title: Enhancing Low-Light Video Summarization via Spatial-Temporal Attention Network Models

Names of Authors: Min-Ji Kim¹, David Vance²

Authors’ Affiliations: ¹School of Electronics and Informatics, Seoul National University, Seoul, South Korea; ²Computer Vision Research Systems, Lumina Tech Labs, Denver, USA

Abstract: Processing video security streams recorded in low-light environments is challenging due to high sensor noise and low color contrast. These artifacts cause standard deep learning object trackers and video summarization models to drop critical frames. This paper introduces a specialized spatial-temporal attention model designed to summarize low-light security video footage accurately. The architecture features an internal feature-enhancement layer that amplifies low-contrast details without over-processing high-frequency sensor noise. A recurrent neural layer calculates regional attention maps across consecutive frames to highlight moving subjects. We trained and evaluated the neural system using a comprehensive low-light security dataset. The model achieved a F1-score optimization level of 0.89, outperforming conventional summarization networks by 14.2%. The total video file processing volume was condensed by 65.0%, preserving critical action sequences. This framework provides an efficient method to scale large-scale urban video analytics infrastructure running under challenging environmental conditions.

Keywords: Video Summarization, Spatial-Temporal Attention, Low-Light Enhancement, Computer Vision, Deep Learning, Security Analytics

Manuscript Timeline: Received: June 02, 2022; Revised: July 08, 2022; Accepted: August 01, 2022; Published: January 06, 2023

International Journal of Computer Science and Technology | Vol. 3, No. 12, December 2022 | pp. 1–8

DOI: 10.46882/2022/IJCST/000215

Article Type: Original Research Paper

Title: Automated Microservice Vulnerability Assessments via Contextual Resource Consumption Profiling

Names of Authors: Klaus Meyer¹, Rachel Green²

Authors’ Affiliations: ¹Institute for Software Systems Security, Munich Technical University, Munich, Germany; ²DevSecOps Research Center, CloudScale Innovations, Austin, USA

Abstract: Monolithic software applications are rapidly transitioning toward cloud-native microservice architectures to achieve higher scalability and deployment flexibility. However, tracking anomalies across hundreds of independent containers introduces significant security and operational complexity. This paper presents an automated security validation framework designed to identify microservice vulnerabilities through non-intrusive container profiling. The system tracks low-level kernel activities, specifically counting system calls alongside memory allocations and CPU usage patterns. It creates a dynamic behavior baseline using an isolation forest algorithm to identify anomalous infrastructure patterns. We tested the tracking mechanism across a continuous integration environment running 80 distinct service modules. The profiling framework successfully identified container escape activities and resource-exhaustion exploits with a 95.4% success rate. The tracking agent operates outside the main application container, maintaining a low resource footprint under 1.8% CPU overhead. This setup ensures continuous software vulnerability assessments without degrading application performance.

Keywords: Microservices, Cloud-Native Security, Container Profiling, System Call Analysis, Automated Detection, DevOps

Manuscript Timeline: Received: May 29, 2022; Revised: July 05, 2022; Accepted: July 29, 2022; Published: December 14, 2022

International Journal of Computer Science and Technology | Vol. 3, No. 11, November 2022 | pp. 1–8

DOI: 10.46882/2022/IJCST/000214

Article Type: Original Research Paper

Title: Lightweight Cross-Layer Congestion Management Protocol for UAV-Assisted Wireless Mesh Networks

Names of Authors: Carlos Silva¹, Fatima Al-Sayed²

Authors’ Affiliations: ¹Department of Mobile Communications, Federal University of Minas Gerais, Belo Horizonte, Brazil; ²Wireless Systems Laboratory, Gulf University for Science, Kuwait City, Kuwait

Abstract: Unmanned Aerial Vehicle (UAV) networks provide critical communication infrastructure during disaster recovery scenarios but frequently suffer from link instability and data congestion. Standard internet transport protocols fail to differentiate between physical channel packet dropouts and buffer congestion states. This research presents a cross-layer congestion routing protocol designed to optimize data transmission across aerial-to-ground communication networks. The protocol links link-layer queue lengths with transport-layer window configurations to estimate network density states accurately. A predictive tracking feature estimates spatial mobility patterns to update network paths before physical signal degradation occurs. We evaluated the communication protocol within a simulated 3D landscape utilizing 15 moving aerial nodes. The tracking results show a 31.5% enhancement in overall data delivery reliability over traditional routing protocols. The average end-to-end packet delivery delay was reduced to 64 ms, stabilizing critical voice and telemetry feeds. This cross-layer coordination prevents throughput degradation caused by rapid node movement in complex aerial environments.

Keywords: Unmanned Aerial Vehicles, Wireless Mesh Networks, Congestion Control, Cross-Layer Optimization, Routing Protocols, Throughput

Manuscript Timeline: Received: May 25, 2022; Revised: July 02, 2022; Accepted: July 26, 2022; Published: November 12, 2022