Learning to match and cluster large high-dimensional data sets for data integration
Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining
Adaptive duplicate detection using learnable string similarity measures
Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining
Learning relational probability trees
Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining
Ensembles of relational classifiers
Knowledge and Information Systems
Scaling record linkage to non-uniform distributed class sizes
PAKDD'08 Proceedings of the 12th Pacific-Asia conference on Advances in knowledge discovery and data mining
| Hi-index | 0.00 |
Relational classifiers use relations between objects to predict the class values. In some cases the relations are explicitly given. In other cases the dataset contains implicit relations, e.g. the relation is hidden inside of noisy attribute values. To apply relational classifiers for this task, the relations have to be extracted. Manually extracting relations by a domain expert is an expensive and time consuming task. In this paper we show how extracting relations in datasets with noisy attribute values can be learned. Our method LRE uses a regression model to learn and predict weighted binary relations. We show that LRE is able to extract both equivalence relations and non-constrained relations. Secondly we show that relational classifiers using relations automatically extracted by LRE achieve comparable classification quality as classifiers using manually labeled relations.