Selected Papers from the Ninth International Conference on Computational Intelligence and Security

Yiu‐ming Cheung, Yuping Wang, Hai‐Lin Liu, Xiaodong LiThe Scientific World JOURNAL 2013 · DOI

TL;DR

This editorial highlights five selected papers from the 2012 International Conference on Computational Intelligence and Security (CIS). These papers showcase innovative applications of computational intelligence techniques to address critical challenges in information security, ranging from data mining for malware detection to robust control systems.

What Problem Does This Paper Address?

This paper is an editorial that introduces a special issue featuring selected works from the 2012 International Conference on Computational Intelligence and Security (CIS). The overarching theme is the application of computational intelligence (CI) models, algorithms, and technologies to solve problems in information security (IS). The editors aim to bring to the forefront research that bridges these two crucial fields, demonstrating how advanced computational methods can enhance security measures and analysis.

Methods

The editorial summarizes five distinct research papers, each employing different methodologies:

  • Association Rule Mining: One paper proposes an evolutionary algorithm for multi-objective association rule mining. This method aims to overcome limitations of traditional approaches by not requiring pre-specified minimum support and confidence thresholds, utilizing attribute index and uniform design.
  • Web Service Composition: Another paper focuses on selecting web services by considering their transactional properties and non-functional Quality of Service (QoS) attributes.
  • Adaptive Neural Control: A third paper presents an adaptive neural control design for uncertain nonlinear systems with time-varying delays. The approach is designed to ensure bounded signals and accurate output tracking.
  • Graph Theory: One study investigates upper bounds for spectral radii in quantum chemistry, specifically deriving a Nordhaus-Gaddum type bound for the Laplacian spectral radius of a graph and its complement.
  • Malware Signature Analysis: The final paper evaluates a static structure approach to malware modeling, leveraging growing malware signature databases. It explores the use of bioinformatics sequence alignment techniques for improved malware classification.

Datasets

Not discussed in the paper. The editorial mentions experiments on "several databases" for the association rule mining paper, but specific details are not provided.

Key Insights and Ideas

  • Bridging CI and IS: The core idea is the synergistic potential of computational intelligence and information security. The selected papers demonstrate concrete examples of how CI can offer novel solutions to complex security problems.
  • Beyond Traditional Thresholds: The association rule mining paper highlights a move away from rigid, user-defined parameters (like minimum support and confidence) towards more adaptive and robust data mining techniques.
  • Leveraging Diverse Domains: The inclusion of papers on control systems, graph theory, and bioinformatics for malware analysis shows the broad applicability of CI principles across different facets of security.
  • Improving Malware Detection: The application of bioinformatics techniques to malware signatures suggests a promising avenue for more accurate and nuanced detection of malicious software.

Key Results

  • The proposed association rule mining algorithm demonstrated excellent performance and significantly reduced the number of comparisons and time consumption.
  • The adaptive neural control scheme was shown to be effective, guaranteeing bounded closed-loop signals and converging output tracking error to a desired neighborhood.
  • A Nordhaus-Gaddum type upper bound was obtained for the sum of Laplacian spectral radii.
  • The static structure approach to malware modeling showed that standard sequence alignment techniques can improve the accuracy of distinguishing between viral and worm signatures when malware signatures are represented as artificial protein sequences.

Limitations and Future Work

  • Limited Specificity: As an editorial, it provides high-level summaries. Detailed limitations and future work for each individual paper are not elaborated upon.
  • General Applicability: The editorial implies that the presented works offer "useful ideas" for the fields of CI and security, suggesting a broader impact beyond the specific problems addressed in each paper. Future work would likely involve further development and application of these techniques to a wider range of security challenges.

Summary updated

Abstract

The 2012 International Conference on Computational Intelligence and Security (CIS) is the ninth one focusing on all areas of two crucial fields in information processing: computational intelligence (CI) and information security (IS). In particular, the CIS Conference provides a platform to explore the potential applications of CI models, algorithms, and technologies to IS.
BibTeX
@article{cheung2013selected,
  author = {Yiu‐ming Cheung and Yuping Wang and Hai‐Lin Liu and Xiaodong Li},
  title = {{Selected Papers from the Ninth International Conference on Computational Intelligence and Security}},
  journal = {The Scientific World JOURNAL},
  year = {2013},
  doi = {10.1155/2013/467321},
}
Figure