The construction of attribute (object)-oriented multi-granularity concept lattices

How to reduce the complexity of lattice construction is an important research topic in formal concept analysis. Based on granularity tree, the relationship between the extent and the intent of the attribute (object)-oriented concept before and after granularity transformation are investigated. Then,...

Full description

Saved in:
Bibliographic Details
Published in:International journal of machine learning and cybernetics 2020-05, Vol.11 (5), p.1017-1032
Main Authors: Shao, Ming-Wen, Lv, Meng-Meng, Li, Ken-Wen, Wang, Chang-Zhong
Format: Article
Language:English
Subjects:
Algorithms
Approximation
Artificial Intelligence
Complex Systems
Complexity
Computational Intelligence
Control
Data mining
Decision making
Engineering
Fuzzy sets
Information systems
Knowledge representation
Lattice theory
Lattices
Mechatronics
Original Article
Pattern Recognition
Robotics
Systems Biology
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:How to reduce the complexity of lattice construction is an important research topic in formal concept analysis. Based on granularity tree, the relationship between the extent and the intent of the attribute (object)-oriented concept before and after granularity transformation are investigated. Then, zoom algorithms for attribute (object)-oriented concept lattices are proposed. Specifically, zoom-in algorithm is applied to change the attribute granularity from coarse-granularity to fine-granularity, and zoom-out algorithm achieves changing the attribute granularity from fine-granularity to coarse-granularity. Zoom algorithms deal with the problems of fast construction of the attribute (object)-oriented multi-granularity concept lattices. By using zoom algorithms, the attribute (object)-oriented concept lattice based on different attribute granularity can be directly generated through the existing attribute (object)-oriented concept lattice. The proposed algorithms not only reduce the computational complexity of concept lattice construction, but also facilitate further data mining and knowledge discovery in formal contexts. Furthermore, the transformation algorithms among three kinds of concept lattice are proposed.
ISSN:1868-8071
1868-808X
DOI:10.1007/s13042-019-00955-0