Efficient mining of association rules using closed itemset lattices
Information Systems
Automatic Structuring of Knowledge Bases by Conceptual Clustering
IEEE Transactions on Knowledge and Data Engineering
Knowledge Discovery from Very Large Databases Using Frequent Concept Lattices
ECML '00 Proceedings of the 11th European Conference on Machine Learning
Learning Classification Rules Using Lattices (Extended Abstract)
ECML '95 Proceedings of the 8th European Conference on Machine Learning
Conceptual Knowledge Discovery in Databases Using Formal Concept Analysis Methods
PKDD '98 Proceedings of the Second European Symposium on Principles of Data Mining and Knowledge Discovery
Formal Concept Analysis for Domain-Specific Document Retrieval Systems
AI '01 Proceedings of the 14th Australian Joint Conference on Artificial Intelligence: Advances in Artificial Intelligence
Representing Large Concept Hierarchies Using Lattice Data Structure
PAKDD '01 Proceedings of the 5th Pacific-Asia Conference on Knowledge Discovery and Data Mining
Integrating Classification and Association Rule Mining: A Concept Lattice Framework
RSFDGrC '99 Proceedings of the 7th International Workshop on New Directions in Rough Sets, Data Mining, and Granular-Soft Computing
Intelligent Structuring and Reducing of Association Rules with Formal Concept Analysis
KI '01 Proceedings of the Joint German/Austrian Conference on AI: Advances in Artificial Intelligence
Predicting protein structural class from closed protein sequences
PAKDD'03 Proceedings of the 7th Pacific-Asia conference on Advances in knowledge discovery and data mining
| Hi-index | 0.00 |
In this paper, we define formal concept mining, a method for generating and evaluating all the pertinent concepts from large transaction databases. We propose a novel efficient formal concept mining algorithm, called Distribution Curve Self-Evaluation (DCSEA). Attempting repeatedly to self-adjust the normal distribution curve to be as close as the symmetry curve, DCSEA automatically identifies all the pertinent concepts by deleting and masking non-pertinent concepts. Instead of using the global support threshold, DCSEA allows users to specify the interestingness of the output concepts by using a more understandable statistic-based threshold, called minimum significance threshold. Such threshold measures the level of significance of the concept extent size (the number of objects) from all the concept extent sizes. Experimental results showed that the proposed algorithm gives high concept retrieval performance, and efficient concept focusing, especially on large databases.