An Approach of Concept Lattice Theory in Data Mining and Its Applications

Concept lattice has been proven to be a very effective tool and architecture for data mining in general. It is widely used for data analysis and knowledge discovery and various concept lattice based approaches are used depending on the type of data. This extended version paper aims at presenting one application of the lattice theory in text mining and another one in image mining.

In the first approach, the notion of lattice theory has been applied by using one of its components mostly used in data mining, the formal concept analysis which has a powerful method, the association rule extraction which helps to find in a database patterns which appear frequently together.

In the second one, the use of the lattice theory for image sets characterization has been shown by using landmarks to enable a machine to automatically classify objects with respect to the image class they belong to. For text mining, association rules discovery, mostly uses formal concept analysis to analyze the relations between patterns which appear at the same time.