Inference Graphs: Combining Natural Deduction and Subsumption Inference in a Concurrent Reasoner
Authors: Daniel Schlegel, Stuart Shapiro
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Evaluation of concurrency characteristics on a combination natural deduction and subsumption reasoning problem has shown linear speedup with the number of processors. We evaluated the performance of IGs in backward inference, as it is the most resource intensive type of inference IGs can perform, and most fully utilizes the scheduling heuristics. To do so we used graphs of chaining andentailments... |
| Researcher Affiliation | Academia | Daniel R. Schlegela and Stuart C. Shapirob a Department of Biomedical Informatics b Department of Computer Science and Engineering University at Buffalo, Buffalo NY, 14260 |
| Pseudocode | No | The paper describes the inference process and message structures, but it does not include any explicitly labeled pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not contain any statement about releasing source code or provide a link to a code repository for the methodology described. |
| Open Datasets | No | The paper describes a self-generated problem instance for evaluation ('graphs of chaining andentailments') rather than referring to a named, publicly accessible dataset with a link or citation. |
| Dataset Splits | No | The paper does not provide specific information about training, validation, or test dataset splits (e.g., percentages, sample counts, or cross-validation details) for reproducibility. |
| Hardware Specification | Yes | Tests were performed on a Dell Poweredge 1950 server with dual quad-core Intel Xeon X5365 processors and 32GB RAM. |
| Software Dependencies | No | The paper mentions various systems and standards like 'CLIF (ISO/IEC 2007)', 'ANALOG', 'Power Loom', but it does not specify any software dependencies with version numbers used for its implementation or experiments. |
| Experiment Setup | Yes | We evaluated the performance of IGs in backward inference... To do so we used graphs of chaining andentailments, meaning for each implication to derive its consequent, all its antecedents had to be true. Each entailment had bf antecedents, where bf is the branching factor, and a single consequent. ... with bf = 2 and d = 7. |