Organizational texts classification using artificial immune recognition systems

Abstract

This paper outlines the use of Artificial Immune Recognition System (AIRS) within the field of text/document classification. Various versions of AIRS including AIRS1, AIRS2, Parallel AIRS and Modified AIRS with Fuzzy KNN are applied to classify the mode of a text's content which is organized for helping users with their organizational tasks. In this regard, 7 major features as inputs with 3 nominal values of Low, Medium, and High are chosen to classify texts into 6 organizational functionality classes. Results of experimentation on a dataset including 540 data show the fact that different versions of AIRS, performs better compared to multi-layer perceptron and radial basis function as simple neural approaches. Due to the high performance of this approach, it is expected to be successfully applicable to a wide range of content mode classification issues in decision support environment.