Learning First-Order Logic Embeddings via Matrix Factorization

Authors: William Yang Wang, William W. Cohen

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In experiments, we demonstrate the effectiveness of reasoning with first-order logic embeddings by comparing with several state-of-the-art baselines on two datasets in the task of knowledge base completion.
Researcher Affiliation Academia William Yang Wang and William W. Cohen School of Computer Science Carnegie Mellon University, Pittsburgh, PA 15213, U.S.A. {yww,wcohen}@cs.cmu.edu
Pseudocode Yes Algorithm 1 A Matrix Factorization Based Algorithm for Learning First-Order Logic Embeddings
Open Source Code No The paper does not provide an explicit statement about the release of its source code or a link to a code repository for the methodology described.
Open Datasets Yes The statistics of the datasets are shown in Table 2. ... Table 2: Statistics of the two publicly available datasets used in the knowledge base completion experiments. Rel.: relations. En.: entities. Word Net 18 40,943 141,442 5,000 5,000 FB15K 1,345 14,951 483,142 50,000 59,071
Dataset Splits Yes Table 2: Statistics of the two publicly available datasets used in the knowledge base completion experiments. ... Word Net 18 40,943 141,442 5,000 5,000 FB15K 1,345 14,951 483,142 50,000 59,071
Hardware Specification No The paper does not specify the hardware (e.g., CPU, GPU models, memory) used for the experiments.
Software Dependencies No The paper mentions using 'stochastic gradient descent (SGD)' and 'fast parallel stochastic gradient descent (FPSG) [Chin et al., 2015]' as optimization approaches, but does not specify particular software libraries or their version numbers.
Experiment Setup Yes Pro PPR s reset parameter is set to 0.1, and the approximation error parameter is set to 1 10 3. ... the latent dimension of first-order logic embeddings is set to 8. ... we vary the latent dimension k for learning first-order logic embeddings. ... We investigate the effects of choosing various loss functions for learning first-order logic embeddings.