Based on the presentations, discussions, and interaction between participants, which occurred during the workshop, the following topics emerged as the upcoming leading research topics in the field of spatio-temporal informatics:
I. New challenges in information extraction and modeling
• Spatiotemporal modeling, especially as it relates to fuzzy and abstract information
• Support for seamless navigation through space and time datasets:
o Continuous updates of databases
o Fully 3-D spaces
o Space/time prediction (e.g. for event monitoring, resource allocation, alert issues)
o Legacy and historical data integration
o Now and then in Google Earth: continuously updating its content, accessing legacy and timely data and information, predicting emerging situations
• Event-driven approaches:
o Event modeling
o Automated attribute recognition
o Event similarity assessment
o Spatiotemporal event mining
o Reasoning
o Risk assessment, etc.
o Integration in a spatiotemporal algebra
• Global monitoring: cross temporal- and spatial-scale analysis
• Mobility, flow, and evolution: from single to composite objects (e.g. cars, pollution front, groups of people, disease risk): modeling, analyzing, and communicating across space, time, and semantic hierarchies
• New representations through the integration of low- and high-level data (e.g. raw image data and interpreted GIS-data)
o expert annotation
o multiple theme-based representations of the same scene as intuitive context descriptions (a critical underpinning to knowledge discovery)
• GeoRealism: At the right space, time, resolution and quality
o Point clouds vs. 3d models; video vs. events; trajectories vs. patterns
o Vector with imagery; text (from wire news) with maps; verbal descriptions with 3d models
o As fast as we need it, and nothing we don’t.
II. Data collection revisited
• Ambient spatial computing: adaptive, multi-modal, sensor-based
• Towards geosensor networks:
o P2P
o Sensors running multiple, isolated services
o In-network, on-the-fly data analysis, interpretation, integration & resource allocation
o An opportunity to reduce the gap between ST Research and DBMSs
• Humans as sensors, text and speech as data: extending temporal and spatial linguistic analysis
• GeoMedia:
o Narrative-to-video and video-to-narrative capabilities
o Text-to-image and image-to-text
• Location-based services revisited: in-situ analytics. Moving visual-analytical power to everyday devices and tasks
• Collaborative use of diverse data sources to track objects and events
• Mind the gap: identifying redundancies and gaps in massive amounts of spatiotemporal datasets
III. Support for cross-discipline discovery using spatiotemporal information
• Decentralized geospatial computing
• Knowledge representation: within and cross-domain (e.g. temporal and event-based modeling for hazards, evacuation, and disaster recovery processes)
• Metadata: Visual/Hierarchical/Dynamic/Self-generated/Integrated/Composite
• Κnowledge discovery tools to build empirical models for domain experts
o Intuitive/visual spatiotemporal queries
o Link between observation, field experiment, laboratory, and theory
o Anomaly detection and causality
o Scenario-based reasoning: support for what-if, counterfactual scenario generation and testing using spatiotemporal information
• Space and time scale harmonization: From atmospheric layers to molecular dynamics
IV. Support for non-expert interaction with spatiotemporal information
• Ambient Spatial Intelligence: Personalized, ubiquitous, location-based services:
o scalability
o privacy
o context- and preference-awareness
• Link to social networks: geo-chatting, geo-twitter (location-aware social informatics – geosocial informatics)
• Ad-hoc, purpose-driven social networking: recognizing common spatiotemporal activities and linking users/carriers to exchange information
o From the classic (transportation services with information shared among neighboring/meeting vehicles) to the more exotic (recognizing patterns of activities and preferences to identify different groups of individuals as they interact with their environment)
• Support for 3-D modeling and interaction of non-professionals with their environment through consumer products (e.g. cameras, cell phones, dashboard-mounted units)
o Geolocating amateur static and video imagery using scene descriptors
o Using amateur data to update databases
o Quality management of volunteered information
o Delivering specialized information to amateur users to aid their navigation in (and interaction with) urban environments
• Data and information delivery onto new modalities (i-pods, phones, cars etc.)
• Dynamic integration/interaction across scales and domains
• New cross-disciplinary paradigms: Rethinking and expanding the chain from society needs to scientific response
• Privacy
NOTE: A Word file containing the preliminary summary of the workshop, based on presentations and discussions held during the event, is available here.