Addconcept: A vertical union algorithm of concept lattices

Assembling same-fields concept lattices into a new concept lattice is considered as a novel research method in parallel constructing concept lattice and ontology merging. Its basic idea is to horizontally or vertically divide a large formal context into a set of smaller sub-contexts that share objects or attributes, and then assemble the corresponding lattices of the sub-contexts. Different from the existing concept lattices union algorithms that all the concepts should be compared, this paper presents a new union approach of concept lattices, in which only the new and updated concepts generated in last insertion are compared. Thus, we can drastically reduce the comparison times of the concepts by traversing the structures of the original lattices. Experiments show that the proposed algorithm can improve efficiency obviously, compared with other concept lattice merging algorithms.