Hierarchical structure analysis of linguistic expressions using concept lattice

Computing with words has proven to be a valuable tool for directly processing linguistic information. However, due to the different states of objects at different times or places, how to dynamically obtain the relations and hierarchical structures of linguistic expressions in different contexts is always a challenge. This paper proposes a computing with linguistic expressions (CWLE) model based on linguistic concept lattices to address this challenge. To handle uncertainty in linguistic expressions, interval type-2 fuzzy sets are first employed to model them, with an initial order established via the centroid mean, enabling flexible adaptation to varied contexts. Second, the linguistic label formal context automates fuzzy set generation, while a fuzzy linguistic-valued lattice is constructed based on the similarity and hierarchical relationships among linguistic expressions. In addition, a hierarchical generation algorithm further captures complex contextual relationships. Finally, comparative analysis demonstrates the CWLE model’s effectiveness in accurately representing the hierarchical structure of linguistic expressions.