International Journal of Computer Science and Technology

International Journal of Computer Science and Technology | Vol. 10, No. 8, August 2008 | pp. 57–64

DOI: 10.46882/2008/IJCST/108017

Research Article

Title: An Architecture-Centric Approach to Evaluating the Reliability of Component-Based Software Systems

Names of Authors: Hans-Werner Schmidt¹, Dieter K. Müller¹

Authors’ Affiliations: ¹Institute of Software Systems, Technical University of Munich, Munich, Germany

Abstract: Modern enterprise applications are increasingly constructed by integrating third-party commercial off-the-shelf (COTS) software components. Evaluating the overall reliability of these systems is challenging due to the hidden internal complexities and opaque failure characteristics of individual black-box components. This paper introduces an architecture-centric analytical framework that uses Discrete-Time Markov Chains (DTMC) to model system operational profiles and compute reliable runtime failure probabilities. The approach maps component interactions, interface dependencies, and transactional frequencies into a coherent execution state graph. We run sensitivity analyses to isolate critical individual components whose failure behaviors present the highest risk to global application stability. Validation experiments conducted on a simulated distributed banking infrastructure showed that the DTMC predictive framework estimated system failure rates within an error margin of ±1.8% compared to empirical load-testing results. This mathematical framework allows software engineers to systematically specify component-level reliability requirements before full integration.

Keywords: Software reliability, component-based software, Markov chains, software architecture, sensitivity analysis, reliability modeling

Manuscript Timeline: Received May 12, 2008; Revised June 25, 2008; Accepted July 19, 2008; Published August 5, 2008

Citation: Schmidt, H. W., & Müller, D. K. (2008). An Architecture-Centric Approach to Evaluating the Reliability of Component-Based Software Systems. International Journal of Computer Science and Technology, 10(8), 57–64.

International Journal of Computer Science and Technology | Vol. 10, No. 7, July 2008 | pp. 49–56

DOI: 10.46882/2008/IJCST/107016

Research Article

Title: Design of an Automated Plant Disease Detection System Using Digital Image Processing and Neural Networks

Names of Authors: Chidi N. Okafor¹, Blessing C. Eze²

Authors’ Affiliations: ¹Department of Electronic Engineering, University of Nigeria, Nsukka, Nigeria; ²Department of Agronomy, Michael Okpara University of Agriculture, Umudike, Nigeria

Abstract: Early detection of agricultural crop diseases is imperative to minimizing yield losses and ensuring global food security. Manual inspection methods are error-prone, highly subjective, and impractical across massive agricultural landscapes. This paper describes the design and testing of an automated leaf disease detection engine tailored for cash crops. The framework uses color segmentation via K-means clustering within the L*a*b* color space to isolate necrotic lesions from healthy leaf tissue. Spatial and statistical properties are extracted using a Gray-Level Co-occurrence Matrix (GLCM), which feeds into a Multi-Layer Perceptron (MLP) neural network trained backpropagation. The system was evaluated on a dataset of 1,200 digital leaf images encompassing healthy, black spot, and rust-infected specimens. The automated diagnostic engine achieved a classification accuracy rate of 94.6% across all test sets. The processing pipeline completed feature extraction and neural inference within an average duration of 720 ± 25 milliseconds per leaf sample, demonstrating high suitability for edge-deployment on mobile diagnostic devices.

Keywords: Image processing, K-means clustering, neural networks, plant disease detection, pattern recognition, precision agriculture

Manuscript Timeline: Received April 9, 2008; Revised May 22, 2008; Accepted June 14, 2008; Published July 7, 2008

Citation: Okafor, C. N., & Eze, C. C. (2008). Design of an Automated Plant Disease Detection System Using Digital Image Processing and Neural Networks. International Journal of Computer Science and Technology, 10(7), 49–56.

International Journal of Computer Science and Technology | Vol. 10, No. 6, June 2008 | pp. 41–48

DOI: 10.46882/2008/IJCST/106015

Research Article

Title: A Scalable Data Mining Infrastructure for Association Rule Discovery on Large Market Basket Datasets

Names of Authors: Thomas E. Keller¹, Sarah J. Fletcher¹

Authors’ Affiliations: ¹Department of Computer Science, University of Illinois at Urbana-Champaign, Urbana, IL, USA

Abstract: Extracting meaningful consumer purchase patterns from massive enterprise data warehouses demands highly efficient data mining frameworks. Standard Apriori-based mining implementations are frequently bottlenecked by the computational cost of iterative database scans and substantial memory requirements for candidate generation. This paper presents a scalable data mining infrastructure featuring an optimized Frequent Pattern Growth (FP-Growth) algorithm that uses a localized horizontal data partitioning paradigm. The architecture distributes the overhead of tree construction across multiple concurrent memory spaces, drastically shrinking global input/output wait times. The system was stress-tested using synthetic transactional datasets scaling up to 10 million individual items. Benchmark logs reveal a 4.2x acceleration in transaction processing speeds relative to traditional single-threaded FP-growth configurations. The infrastructure maintained linear scalability, keeping peak RAM consumption stable at 1.2 GB ± 0.1 GB during high-density itemset passes, rendering it highly viable for real-time retail decision-support systems.

Keywords: Data mining, association rules, market basket analysis, FP-growth, scalable architecture, parallel processing

Manuscript Timeline: Received March 3, 2008; Revised April 15, 2008; Accepted May 12, 2008; Published June 4, 2008

Citation: Keller, T. E., & Fletcher, S. J. (2008). A Scalable Data Mining Infrastructure for Association Rule Discovery on Large Market Basket Datasets. International Journal of Computer Science and Technology, 10(6), 41–48.

International Journal of Computer Science and Technology | Vol. 10, No. 5, May 2008 | pp. 33–40

DOI: 10.46882/2008/IJCST/105014

Research Article

Title: Modeling Object-Oriented Software Maintenance Costs Using Adaptive Neuro-Fuzzy Inference Systems

Names of Authors: Hasan M. Tarawneh¹

Authors’ Affiliations: ¹Department of Software Engineering, Al-Balqa Applied University, Salt, Jordan

Abstract: Accurately projecting software maintenance efforts during the early stages of the software development life cycle remains notoriously difficult due to the non-linear interactions of structural codebase properties. This study develops a predictive model for object-oriented software maintenance effort utilizing an Adaptive Neuro-Fuzzy Inference System (ANFIS). The framework utilizes Chidamber and Kemerer (C&K) metrics—specifically Depth of Inheritance Tree (DIT), Coupling Between Object Classes (CBO), and Weighted Methods per Class (WMC)—as key inputs to map out architectural complexity. The ANFIS network was trained and validated using empirical data extracted from 24 historical open-source Java projects. The system’s predictive accuracy was evaluated using the Mean Absolute Relative Error (MARE). Experimental analysis showed that the ANFIS model achieved a low MARE of 11.4%, significantly outperforming traditional multiple linear regression models (which yielded a MARE of 26.8%) and standalone artificial neural networks (yielding a MARE of 18.2%). The model offers software project managers a precise tool for optimizing human resource allocations during post-deployment phases.

Keywords: Software maintenance, ANFIS, Chidamber and Kemerer metrics, software metrics, neuro-fuzzy systems, cost estimation

Manuscript Timeline: Received February 14, 2008; Revised March 28, 2008; Accepted April 18, 2008; Published May 10, 2008

Citation: Tarawneh, H. M. (2008). Modeling Object-Oriented Software Maintenance Costs Using Adaptive Neuro-Fuzzy Inference Systems. International Journal of Computer Science and Technology, 10(5), 33–40.

International Journal of Computer Science and Technology | Vol. 10, No. 4, April 2008 | pp. 25–32

DOI: 10.46882/2008/IJCST/104013

Research Article

Title: A Robust Digital Image Watermarking Scheme in the Wavelet Domain for Copyright Protection

Names of Authors: Carlos M. Ortega¹, Sofia L. Gomez¹

Authors’ Affiliations: ¹Department of Signal Processing, Polytechnic University of Catalonia, Barcelona, Spain

Abstract: Unauthorized duplication and redistribution of digital multimedia assets highlight the pressing requirement for resilient copyright protection technologies. This paper introduces an imperceptible and robust digital image watermarking scheme executed within the Discrete Wavelet Transform (DWT) domain, augmented by Singular Value Decomposition (SVD). The algorithm embeds a binary watermark logo into the singular values of selected mid-frequency sub-bands (LH and HL) of the host image. To enhance security and prevent geometric tampering bypasses, the watermark undergoes chaotic encryption via an Arnold transform prior to embedding. Experimental testing across standard greyscale images (including Lena, Barbara, and Cameraman) demonstrated high visual fidelity, achieving an average Peak Signal-to-Noise Ratio (PSNR) of 44.2 dB ± 0.5 dB. The scheme exhibits strong resilience against common signal processing operations; the Normalized Correlation (NC) coefficients of extracted watermarks remained above 0.92 when subjected to JPEG compression down to a 30% quality factor, 3x3 median filtering, and Gaussian noise additions up to a variance of 0.01.

Keywords: Digital watermarking, discrete wavelet transform, singular value decomposition, copyright protection, Arnold transform, robustness analysis

Manuscript Timeline: Received January 10, 2008; Revised February 24, 2008; Accepted March 19, 2008; Published April 6, 2008

Citation: Ortega, C. M., & Gomez, S. L. (2008). A Robust Digital Image Watermarking Scheme in the Wavelet Domain for Copyright Protection. International Journal of Computer Science and Technology, 10(4), 25–32.

International Journal of Computer Science and Technology | Vol. 10, No. 3, March 2008 | pp. 17–24

DOI: 10.46882/2008/IJCST/103012

Research Article

Title: Dynamic Spectrum Allocation in Cognitive Radio Networks Using Evolutionary Game Theory

Names of Authors: Kofi A. Mensah¹, Osei W. Boateng²

Authors’ Affiliations: ¹Department of Telecommunication Engineering, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana; ²Advanced Communications Laboratory, Council for Scientific and Industrial Research, Accra, Ghana

Abstract: Spectrum scarcity represents a major barrier to accommodating the accelerating demands of contemporary wireless networks. Cognitive Radio (CR) technology offers an opportunistic solution, yet optimizing spectrum allocation among secondary users without interfering with primary licensees remains an open challenge. This paper models the dynamic spectrum sharing problem as a non-cooperative multi-player game, applying evolutionary game theory to derive stable network equilibria. We propose an adaptive replicator dynamics algorithm that drives secondary user nodes to dynamically alter their channel selections based on localized utility feedback, which encapsulates bandwidth availability, pricing, and signal-to-interference-plus-noise ratios (SINR). Network simulations demonstrate that the algorithm consistently converges to a Nash Equilibrium within 45 iterations across highly fluid topologies. Secondary user throughput increased by 31.8% compared to static allocation models, while maintaining primary user interference below the threshold limit of -110 dBm. The network achieved an aggregate spectrum utilization efficiency of 89.6% ± 0.7%, indicating high structural robustness for heterogeneous wireless communications.

Keywords: Cognitive radio, spectrum allocation, evolutionary game theory, Nash equilibrium, replicator dynamics, interference mitigation

Manuscript Timeline: Received December 5, 2007; Revised January 20, 2008; Accepted February 15, 2008; Published March 8, 2008

Citation: Mensah, K. A., & Boateng, O. W. (2008). Dynamic Spectrum Allocation in Cognitive Radio Networks Using Evolutionary Game Theory. International Journal of Computer Science and Technology, 10(3), 17–24.