Datasets
Standard Dataset
DATA_AsbestosReveal
- Citation Author(s):
- Submitted by:
- Thamer MECHARNIA
- Last updated:
- Wed, 03/22/2023 - 10:58
- DOI:
- 10.21227/3bcv-6t97
- Data Format:
- License:
60 ViewsAbstract
This data represent a knowledge graph (KG) that has been populated using a set of diagnostic documents provided by the CSTB (Scientific and technical center for building). This KG contains 51970 triples that describe 2998 product instances, 341 locations, 214 structures and 94 buildings. The construction year of those buildings varies between 1948 and 1997. We have 1525 products that contain asbestos and 1473 products are asbestos-free. To evaluate our approaches, CRA-Miner and the hybrid approache, we divided the KG data into 3 tiers, and we performed cross-validation.
In our study, we conducted a cross-validation procedure using the KG (Knowledge Graph) data, which involved dividing the data into three tiers. We conducted three tests, where each test involved learning rules from two tiers and testing them on the remaining third tier.
In each directory, we stored the files related to a specific test. Within each directory, there were two sub-folders named 'learningSet' that contained the data for the two tiers used for training, along with a list of the learned rules. The other sub-folder, named 'testingSet,' contained the data from the remaining tier and the corresponding results obtained from testing the learned rules.
