Learning First-Order Rules with Differentiable Logic Program Semantics
Authors: Kun Gao, Katsumi Inoue, Yongzhi Cao, Hanpin Wang
IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We demonstrate that DFOL can perform on several standard ILP datasets, knowledge bases, and probabilistic relation facts and outperform several well-known differentiable ILP models. Experimental results indicate that DFOL is a precise, robust, scalable, and computationally cheap differentiable ILP model. |
| Researcher Affiliation | Collaboration | Kun Gao1 , Katsumi Inoue2 , Yongzhi Cao1 and Hanpin Wang3,1 1Key Laboratory of High Confidence Software Technologies (MOE), School of Computer Science, Peking University 2National Institute of Informatics 3School of Computer Science and Cyber Engineering, Guangzhou University |
| Pseudocode | Yes | Algorithm 1 The propositionalization method in DFOL |
| Open Source Code | No | The paper does not include an unambiguous statement or direct link where the authors state they are releasing the source code for the methodology described in this paper. |
| Open Datasets | Yes | We present the results of DFOL on 20 classifications of ILP datasets [Evans and Grefenstette, 2018]. |
| Dataset Splits | Yes | For the Nations and UMLS datasets, we divide each dataset into 80% training facts, 10% development facts, and 10% test facts. |
| Hardware Specification | Yes | All experiments are executed on 24GB of memory and an 8-core Intel i7-6700 CPU. |
| Software Dependencies | No | The paper mentions general software components like "NNs" and implies a programming environment but does not specify particular library names with version numbers (e.g., PyTorch 1.9, TensorFlow 2.x) or specific solver versions. |
| Experiment Setup | Yes | vo = e m k=1(ϕ(MP [k, ] v T i 1)), LI = H( vo, vo), where the activation function ϕ is defined in Equation (2b). The function H denotes the binary cross-entropy function. (...) We use the Adam algorithm [Kingma and Ba, 2015] to minimize the final loss. (...) We set τs = 1. (...) The standard deviation σ ranges from 0.5 to 3 with step 0.5. (...) mislabeled with the mutation rate µ ranging from 0.05 to 1 with step 0.05. |