Fuzzy rule base interpolation based on semantic revision

Sometimes it is no possible to have a full dense rule base as there are gaps in the information. Further, it is often necessary to deal with sparse rule base to reduce the size and the inference I control time. In such sparse rule bases the classic algorithms like the CRI of Zadeh and the Mamdani-method do not function for observation hitting into the gaps between rules. A linear fuzzy rule interpolation technique (KH-interpolation) has been introduced, that is suitable for dealing with sparse bases, however, this method often results into conclusions which are not directly interpretable. In this paper an interpolation technique is proposed that is based on the interpolation of the semantic and interrelation of rules. This method garantiees the direct interpretability of the conclusion. The comparison of two (KH and BK) and the new interpolation method will also be discussed.