Querying Log Data with Metric Temporal Logic
Authors: Sebastian Brandt, Elem Güzel Kalaycı, Vladislav Ryzhikov, Guohui Xiao, Michael Zakharyaschev
JAIR 2018 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We demonstrate by two real-world use cases that nonrecursive datalog MTL programs can express complex temporal concepts from typical user queries and thereby facilitate access to temporal log data. Our experiments with Siemens turbine data and Meso West weather data show that datalog MTL ontology-mediated queries are efficient and scale on large datasets. |
| Researcher Affiliation | Collaboration | Sebastian Brandt from Siemens CT (an industrial company) and Elem Güzel Kalaycı, Guohui Xiao from Free University of Bozen-Bolzano, Vladislav Ryzhikov, and Michael Zakharyaschev from Birkbeck, University of London (academic institutions) indicate a collaboration between industry and academia. |
| Pseudocode | Yes | The paper includes a structured algorithm in Figure 3 titled "The algorithm for evaluating datalognr MTL queries in SQL.", presenting functions like ans, views, view, projects, and join-cond with detailed steps. |
| Open Source Code | No | The paper states: "Based on these encouraging results, we plan to include our temporal OBDA framework into the Ontop platform... visit http://ontop.inf.unibz.it/ for more information on Ontop." The phrase "plan to include" indicates that the code for the described framework is not yet openly available or provided in this paper, but is a future intention. |
| Open Datasets | Yes | The Meso West (http://mesowest.utah.edu/) project makes publicly available historical records of the weather stations across the US showing such parameters of meteorological conditions as temperature, wind speed and direction, amount of precipitation, etc. |
| Dataset Splits | No | The paper describes how the data was scaled and selected for experiments (e.g., "We replicated this sample to imitate the data for one turbine over 10 different periods..." and "we took the New York state data for the 10 continuous periods between 2005 and 2014..."), but it does not specify any training, testing, or validation splits for model evaluation. |
| Hardware Specification | Yes | We ran the experiments on an HP Proliant server with 2 Intel Xeon X5690 Processors (with 12 logical cores at 3.47GHz each), 106GB of RAM and five 1TB 15K RPM HD. |
| Software Dependencies | Yes | We used both Postgre SQL 9.6 and the SQL interface (Armbrust, Xin, Lian, Huai, Y., Liu, D., Bradley, J. K., Meng, X., Kaftan, T., Franklin, M. J., Ghodsi, A., & Zaharia, M., 2015) of Apache Spark 2.1.0. |
| Experiment Setup | Yes | We ran all the queries with a timeout of 30 minutes. |