Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Intelligent System for Urban Emergency Management during Large-Scale Disaster
Authors: Xuan Song, Quanshi Zhang, Yoshihide Sekimoto, Ryosuke Shibasaki
AAAI 2014 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results are presented in Section 5. We evaluated our system from two aspects: performance of mobility simulation for population flow and performance of destination prediction for individual person. |
| Researcher Affiliation | Academia | Center for Spatial Information Science, The University of Tokyo |
| Pseudocode | Yes | Algorithm 1: Expected Action Frequency Calculation |
| Open Source Code | No | No explicit statement about providing open-source code for the methodology described in this paper. |
| Open Datasets | No | The proposed system stores and manages GPS records of approximately 1.6 million anonymized users throughout Japan from 1 August 2010 to 31 July 2011, which contains approximately 9.2 billion GPS records, more than 600GB csv files. No public access information provided. |
| Dataset Splits | Yes | To evaluate the simulation results of population flow, we performed K-fold cross-validation. The whole disaster data were randomly partitioned into three sub-samples: one sample was used as validation data while the other two were used as training data. |
| Hardware Specification | No | No specific hardware details (e.g., GPU/CPU models, memory) used for running experiments are provided. |
| Software Dependencies | No | No specific software dependencies with version numbers are mentioned. |
| Experiment Setup | Yes | We set cell length as 1km, and manually labeled the region type in mobility graph. We randomly selected 80% trajectories of the disaster data (18 hours after the earthquake) to train the inference model, and used the remaining 20% data for testing and evaluation. |