In this repository, we present the SPARQL* queries developed for "Spatiotemporal Data Modelling for Epidemiological Research in Hospitals" 10.1109/JBHI.2024.3417224 and the dataset in RDF* generated for testing them.
Below, we present some other related repositories that may be of interest to you:
- HospitalGeneratorRDF: It is also linked to 10.1109/JBHI.2024.3417224
We have designed and implemented in SPARQL* a set of 6 queries based on patients' movements within a hospital that are meant to assist in several epidemiological research tasks. Each query is in a separate file, where the values of the parameters are the ones used in the experiments presented in 10.1109/JBHI.2024.3417224. The objectives of each query are the following:
- Q1: Detection of an outbreak at a Service
- Q2: Detection of an outbreak at a Location
- Q3: Study of the spread of an outbreak from a patient via analysis of contacts
- Q4: Study of the spread of an outbreak from a set of patients via analysis of contacts
- Q5: Investigation of sources of contagion via analysis of contacts
- Q6: Discovery of the index patient
We provide the dataset used for the experiments presented in 10.1109/JBHI.2024.3417224 This is an RDF* knowledge graph that follows the spatiotemporal data model presented in the paper. It is a synthetic dataset whose data have been generated using the H-Outbreak simulation model. The values for the parameters of H-Outbreak used to create the dataset are the following:
- n_patients: 0.7
- steps: 462 (462×8 = 3696 Hours → 3696/24 = 154 Days → 154/7 = 22 Weeks)
- population: 170000
- step_time: 8 (hours)
- init_exposed: 1
- init_infected: 0
- arrival_rate: 17.55
- prob_arrival_ER: 0.7
- arrival_state_colonized: 0.076
- arrival_state_S: 0.9973429
- arrival_state_I: 0.001563
- arrival_state_NS: 0.0010941
- prob_p-env_min: 0.14
- prob_p-env_max: 0.9
- prob_p-env_mean: 0.52
- prob_env-p_min: 0.3262
- prob_env-p_max: 0.5437
- prob_env-p_mean: 0.435
- prob_pe_min: 0.18
- prob_pe_max: 0.3
- prob_pe_mean: 0.24
- incubation_time_min: 48
- incubation_time_max: 72
- prob_quick_recov_min: 0.0
- prob_quick_recov_max: 0.23
- prob_quick_recov_mean: 0.115
- prob_long_recov_min: 0.5985
- prob_long_recov_max: 0.9975
- prob_long_recov_mean: 0.7981
- treatment_days_min: 5
- treatment_days_max: 15
- treatment_days_mean: 10
- prob_death: 0.069
- max_patients_rx: 3
- max_patients_qx: 2
- min_steps_rx: 10
- min_steps_qx: 30
- max_ward_movements: 2
- max_steps_er_icu: 3
- max_movements_room: 5
- occupancy_icu: 0.46
To transform the output from H-Outbreak in an RDF* knowledge graph, we have used HospitalGeneratorRDF. The file Dataset/LogicZones.ttl has been created ad hoc and contains both the nodes and the edges related to the class LogicZone. The values for the parameters used to create the hospital and the temporal data are the following:
- index: 1600
- huPerService: 3
- nFloors: 5
- huPerFloor: 6
- nRows: 3
- nColumns: 4
- startDateTime: 01/01/2023 08:00:00
- optionFloorUH: None