An approach to modeling short messages in spatio- temporal - PowerPoint PPT Presentation

An approach to modeling short messages in spatio- temporal networks Amosse EDOUARD, PhD student Nhan LE-THAN, Supervisor

Outline 1. Concept & Objective 2. The Radio Social Platform 3. Realisation 4. Contribution 5. Perspectives 6. Current Works 2

Short messages — People prefere to communicate with short messages — Time constraints — Easy to produce & to share — Technicals limitations/restrictions of mobile sensors 3

Motivation « This traffic jam bothers me! » « I am on Route des Lucioles at 6:00 P .M and there is a traffic jam which bothers me » 4

Objective — Semantic enrichment — Spatial properties & thematic — Platform for creating, managing and sharing spatio- temporal information. 5

Radio Social Thematic Listener Reporter Radio social Broadcas Redaction ting 6

Use case Moderate Traffic Moderate traffic e-reporters e-redaction Traffic Event e-diffusion e-listeners Moderate traffic on Route des Colles Moderate traffic on Route des Colles 7

Message structures — Defined by the thematic — Qualificative : tags, symbols — Descriptive 8

BRIEF Message modeling & context enrichment T R — Message Semantic context — RS metadata — Thematic M — Reporter — Space — Time Spatio-temporal context T S 9

Context infering — Top down approach à Non trivial problem — Bottom up approach à The approach we use è We aim to enrich the information context regard to user context 10

Spatial Representation — Geographical space — Set of dynamic and static objects 11

Spatial modeling — Dynamic spatial entities are modelized regard to static entities — Semantic and GIS infrastructures — Google Maps & Places services — Geonames ontology 12

Exemple Geonames & Google Places 43.6161871, 7.0677087 { "long_name" : "Route des Lucioles", "short_name" : "Route des Lucioles", "types" : [ "route" ] }, :geo a :Feature shortName : "Route des Lucioles"; name : "Route des Lucioles"; wgs84_pos:lat "43.6161871"; wgs84_pos:long "7.0677087"; 13

Spatial layers overlapping 14

Time modelling — We used the OWL Time — Instant for messages :messageTi me a :Instant; :inXSDTime : 2014-03-20T10:30:00-5:00 ; — Interval for event :eventTime a :Interval; :hasBeginning: :eventStart ; :eventStart a :Instant; :inXSDTime 2014-03-20T10:30:00-5:00 ; 15

The Ontology — 9 classes — 13 ObjectProperty — 6 DatatypeProperty — Using existing ontologies — OWL Time — GeoNames & WGS84 — FOAF 16

Radio Sociale Ontology 17

Technical architecture JENA BEAN Entity Manager JENA Model SPARQL EJB Container Jax-B SPARQL Endpoint REST Services / Data Model SPARQL Virtuoso 18

Other use case — Epidemic symptom report — The frequency : RASE — Annotation structure — Disease : Flu — Symptom : cough, rheum — Metakeys : reliability, certitude 19

RASE : Use case Fiever Flux Antibes Valbonne Flu cough 12/12/13 Fiever, cough 12/12/13 Flu symptom at Biot Emergence of a flu epidemic in Nice 20

Contributions — Short messages modeling — Spatial mobile Gentities modeling — The radio sociale ontology — Generic platform radio social platform 21

Research keys and limitations — Modeling thematic using existing ontologies (DBPedia) — Ontology-based user interfaces — Spatial reasoning — What could be the incidence of an event on a region to others 22

Current works — Qualitative spatial representation — Spatial reasonning 23

Thank you ! 24