Tong Xiao-Jun

Harbin Institute of Technology at Weihai, Wei-hai-shih, Shandong Sheng, China

Are you Tong Xiao-Jun?

Claim your profile

Publications (2)0 Total impact

  • Tong Xiao-jun · Cui Ming-gen
    [Show abstract] [Hide abstract]
    ABSTRACT: Chaos defined in general in quantitative characteristics is not a strict definition .In this paper two new chaos functions are proposed and theoretical proof about chaos is finished based on a strict Devaney definition. Using composite chaotic system character, a composite chaos sequences cipher algorithm is designed, which map have uniform distribution function and delta-like self-correlation, and can produce balance nature series to 0-1 after quantified. The unchanged probability of composite chaos makes ciphertext has a good random. As the sensitivity to initial condition and random nature to iterative function choice, key and plaintext and ciphertext form complex and sensitive nonlinear relations, and the correlation is very small. The leaking of ciphertext to key and plaintext information can effectively be prevented. The experimental result show that the system has a very high safety and the key space is expanded.
    No preview · Conference Paper · Oct 2007
  • Source
    Tong Xiao-Jun · Cui Ming-Gen · Song Guo-Long
    [Show abstract] [Hide abstract]
    ABSTRACT: Automatic text classifying is an import application of the information processing technology. This paper introduces the key techniques of Chinese text categorization such as text preprocessing, feature selection, feature representation, training and classifying algorithm, especially analyses the current most important several feature selection methods with emphasis. A Chinese text classifier based on KNN algorithm was developed. The system can preferably implement Chinese automatic text categorization and has a higher quality. We also use this classifier to compare several feature selection methods. In the end, we utilize the experiment results to prove the importance role of feature selection in text categorization.
    Preview · Conference Paper · Oct 2007