Query Answering with Guarded Existential Rules under Stable Model Semantics
Authors: Hai Wan, Guohui Xiao, Chenglin Wang, Xianqiao Liu, Junhong Chen, Zhe Wang3017-3024
AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We have implemented our approach in a prototype system. The evaluation over a set of benchmarks shows encouraging results. Finally, we develop a prototype and conduct experiments on a set of benchmarks. The results confirm that the our approach is scalable for query answering with GNTGDs. |
| Researcher Affiliation | Academia | 1School of Data and Computer Science, Sun Yat-sen University, Guangzhou 510006, P.R.China 2KRDB Research Centre, Free University of Bozen-Bolzano, Piazza Domenicani 3, 39100 Bolzano, Italy 3School of Information and Communication Technology, Griffith University, Brisbane, QLD 4111, Australia |
| Pseudocode | No | The paper contains theoretical discussions, lemmas, and theorems, but no pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | Yes | We implemented a prototype system in Python for query answering with GNTGDs2. To the best of our knowledge, this is the first system with such functionality. The code and data to reproduce the experiments are in the online appendix. 2Code and datasets. https://github.com/sysulic/GNTGDs |
| Open Datasets | Yes | We considered three GNTGDs: LUBM3, Geo Concepts4, and Vicodi5 and two (multi-)linear TGDs: DEEP100(-200)6 as benchmarks... 3LUBM. http://swat.cse.lehigh.edu/projects/lubm/ 4Geo Concepts. http://www.kr.tuwien.ac.at/research/projects/myits/ 5Vicodi. http://www.vicodi.org 6DEEP-100(-200). https://github.com/dbunibas/chasebench |
| Dataset Splits | No | The paper provides dataset sizes but does not specify training, validation, or test splits. For example, Table 1 shows |D| (database size) values like 100, 122, 165, 1000, but no explicit split percentages or counts for training/validation/test sets. |
| Hardware Specification | Yes | All experiments run in 64-bit Linux on a machine with 2.10GHz Intel Xeon and 128G 1333 MHz memory. |
| Software Dependencies | Yes | We use the ASP solver clingo-4.4.0. |
| Experiment Setup | No | The paper describes the overall system (MGIF, MGCF modes) and the benchmarks used but does not provide specific details such as hyperparameters (e.g., learning rates, batch sizes, number of epochs) or other system-level training settings. |