Dynamic Hybrid Type Mining in an Intelligent e-Government Model

This paper presents a new methodology for intelligent e- Government (eGov for short). At first, we investigate the relationship between administration and a citizen in sociology, and propose a model of intelligent eGov corresponding to new society. Furthermore, the intelligent eGov model can be divided into two sub-models with respect to the two civic viewpoints: civic centric service and civilian collaboration, respectively. Based on the intelligent eGov model, we developed the hybrid type mining as a new methodology for classifying civic contents of a question into a target category. In this processing, he/she does not need to understand the meaning of a question sentence. After decomposing into the feature words according to a text, a sentence, and a word, text mining is carried out. Thus, the efficiency of selection of the history for a classification can be improved. The other important contribution in this study is that we provide an approach for automatic generation of a service chain, which is in the process for generating an answer. Finally, we describe the method of acquiring from the question data which already accumulated useful knowledge in the activity of administration.