Optimizing the Method for Building an Extended Linguistic Hierarchy

In those problems dealing with linguistic information and multiple sources of information may happen that the sources involved have different degree of knowledge about the problem and could be suitable and necessary the use of different linguistic term sets with different granularity defining a multi-granular linguistic context. Different approaches have been presented to deal with this type of context, being the linguistic hierarchies [1] an approach quite interesting due to its accuracy in computational model but with a strong limitation about the term sets that can be used. We presented an extension of the linguistic hierarchies [2] to deal any linguistic term set in a precise way. This new approach presents initially a drawback, it needs a term set with a very high granularity, implying complexity in computing with words processes. Therefore, we propose an optimization to building an extended linguistic hierarchy in order to decrease the granularity of such a term set.