On Concept Lattices as Information Channels
Francisco José Valverde Albacete, Carmen Peláez-Moreno, Anselmo Peñas · Concept Lattices and their Applications · 2014
This paper explores the idea that a concept lattice is an in- formation channel between objects and attributes. For this purpose we study the behaviour of incidences in L-formal contexts where L is the range of an information-theoretic entropy function. Examples of such data abound in machine learning and data mining, e.g. confusion matri- ces of multi-class classifiers or document-term matrices. We use a well- motivated information-theoretic heuristic, the maximization of mutual information, that in our conclusions provides a flavour of feature selection providing and information-theory explanation of an established practice in Data Mining, Natural Language Processing and Information Retrieval applications, viz. stop-wording and frequency thresholding. We also in- troduce a post-clustering class identification in the presence of confusions and a flavour of term selection for a multi-label document classification task.