Practical TBox Abduction Based on Justification Patterns
Authors: Jianfeng Du, Hai Wan, Huaguan Ma
AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results demonstrate that combining the proposed methods is able to achieve a practical approach to TBox abduction. To show that the proposed methods can constitute a practical approach to TBox abduction, we conduct experiments on ten coherent TBoxes with various complexity. |
| Researcher Affiliation | Academia | Jianfeng Du,1,2 Hai Wan,3,4 Huaguan Ma3 1Collaborative Innovation Center for 21st-century Maritime Silk Road Studies, Guangdong University of Foreign Studies, Guangzhou 510420, P.R.China 2Cisco School of Informatics, Guangdong University of Foreign Studies, Guangzhou 510006, P.R.China jfdu@gdufs.edu.cn 3School of Data and Computer Science, Sun Yat-sen University, Guangzhou 510006, P.R.China 4Guangdong Key Laboratory of Big Data Analysis and Processing, Guangzhou 510006, P.R.China wanhai@mail.sysu.edu.cn |
| Pseudocode | Yes | Algorithm 1 Compute Justification Patterns(T ) ... Function Find Justification(Su, Sc, A B) |
| Open Source Code | Yes | The implemented system and test TBoxes are available at http://www.dataminingcenter.net/tboxabd/. |
| Open Datasets | Yes | We collected ten coherent TBoxes. Some of their statistics are reported in Table 1. ... LUBM (Guo, Pan, and Heflin 2005) TBox. ... UOBM (Ma et al. 2006) TBoxes. ... The next four were collected from http://protegewiki.stanford.edu/wiki/Protege Ontology Library. ... The last three TBoxes are large TBoxes about medical extensions and were collected from http://www.opengalen.org/download/opengalen8-owl-sources.zip. |
| Dataset Splits | No | The paper does not explicitly provide training/validation/test dataset splits in the conventional machine learning sense for its TBoxes. While it describes generating observations for testing, it does not delineate explicit 'training' or 'validation' sets from the TBoxes themselves. |
| Hardware Specification | Yes | All experiments were conducted on a laptop with Intel Dual-Core 2.60GHz CPU and 8GB RAM, running Windows 7 (64 bit) with the maximum Java heap size set to 8GB. |
| Software Dependencies | Yes | We implemented both the method for computing all justification patterns and the method for computing all JP-based explanations in Java, using the Pellet (Sirin et al. 2007) API (version 2.3.1) to discover and check entailments, and using My SQL to manage justification patterns. |
| Experiment Setup | Yes | We set a time limit of 10000 seconds for computing JP-based explanations whose cardinalities are k (where k {1, 2}) for all 50 generated observations in a test TBox. |