Deontic Sensors
Authors: Julian Padget, Marina De Vos, Charlie Ann Page
IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We exercise our implementation of the ROA architecture pattern (Section 4) with a prototypical use case that examines the service performance under load by varying two parameters: the number of queries to be processed and the number of (client) agents. We run two synthetic workload tests between a laptop client and a server across a wide-area network. The first test sends the service a batch of queries, while the second uses a number of synthetic agents to generate batches of queries. The first stresses throughput, while the second aims to mimic workload in a typical system. Timings are averaged over 100 runs and are the difference between the sending of the first query and the receipt of the last result, in each case. (Section 5, Evaluation) |
| Researcher Affiliation | Academia | Julian Padget, Marina De Vos and Charlie Ann Page Department of Computer Science, University of Bath, United Kingdom {j.a.padget,m.d.vos,c.a.page}@bath.ac.uk |
| Pseudocode | Yes | Listing 2: Alice s plan in Jason (Section 4) provides a structured, code-like representation of an agent's plan, serving the function of an algorithm or pseudocode. |
| Open Source Code | Yes | Inst AL-REST provides Inst AL as a service with a RESTful interface... Code available at https://github.com/instsuite/instal-rest/ (Section 4, footnote 3) |
| Open Datasets | No | The paper states, "Instal-REST itself is validated using the Inst AL test suite (comprising some 1300 tests)." (Section 5). However, no concrete access information (e.g., specific link, DOI, repository name, or formal citation with authors/year for public access) is provided for this test suite. |
| Dataset Splits | No | No specific dataset split information (e.g., exact percentages, sample counts, or citations to predefined splits) for training, validation, or testing is provided. While a "test suite" is mentioned, details on its split or how it's partitioned are absent. |
| Hardware Specification | No | The paper vaguely mentions: "We run two synthetic workload tests between a laptop client and a server across a wide-area network." (Section 5). This description lacks specific hardware details such as CPU/GPU models, memory specifications, or server configurations. |
| Software Dependencies | No | The paper lists various software components such as Jason, Inst AL, Ans Prolog, Docker, Celery, Rabbit MQ, Gunicorn, Postgres, and clingo, but it does not specify version numbers for these dependencies, which is required for reproducible software dependency information. |
| Experiment Setup | Yes | The paper details specific parameters for its synthetic workload tests: "varying two parameters: the number of queries to be processed and the number of (client) agents." and specifies batch sizes for queries (2, 25, 50, 100, 250) and number of agents (1, 5, 10, 25) with associated queries (2, 5, 10). It also mentions using "12 (Celery) workers." (Section 5, Evaluation) |