Discovering action rules from action reducts and their application to failed states data
Eliza Kielak , Zbigniew W. Raś
AbstractA new strategy for extracting generalized action rules is presented in this paper. The notion of an action rule was introduced by Ras and Wieczorkowska in . Based on their construction methods, action rules discovery can be divided into two types: rule-based and object-based. In rule-based approach, action rules are built from certain pairs of classification rules. In object-based approach, action rules are extracted directly from a dataset. They provide an insight of how values of some attributes need to be changed so certain objects can be shifted to a desirable group. The concept of a generalized action rule was introduced in . In this paper, we introduce the notion of object-based reducts and set-based reducts, and propose a strategy for discovering generalized action rules and then action rules directly from them. An interesting strategy for discovering generalized action rules using so called summaries was given by Tzacheva . For the application domain we have chosen publicly available Failed States Data . A failed state is a state perceived as having failed at some of the basic conditions and responsibilities of a sovereign government .
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.