https://github.com/UFOKN/Knowledge-Graph
Revision 6f9d6f2093391745ecf8a24a58f5badbb5ea69dd authored by Adam Shepherd on 31 January 2020, 17:17:27 UTC, committed by GitHub on 31 January 2020, 17:17:27 UTC
1 parent 528f3c6
Tip revision: 6f9d6f2093391745ecf8a24a58f5badbb5ea69dd authored by Adam Shepherd on 31 January 2020, 17:17:27 UTC
Create CompetencyQuestions.md
Create CompetencyQuestions.md
Tip revision: 6f9d6f2
README.md
# Knowledge-Graph
Keeps track of concepts, models, and axioms for the data
## Pitch
- [ ] Computational Capability
- [ ] Persona development
- [ ] How Personnas innteract w/ system
## Prototypes
- [ ] Notify me of potential flood (Nervous Norman). Is this paramterized? given a certain radius from a centroid?
- [ ] Where is it flooded?
- [ ] 1-hr predictions
- [ ] model run
- [ ] Affected transportation routes (Powerful Pete/Proactive Paula) - i.e. From *here* to *there* what *route* is best for a given *vehicle type*?
- [ ] Responding to event - should I take a boat?
## Conceptual Model
GOAL: to communicate the knowledge aspect of the network
### Concepts
- [ ] Vector files - graph stores Digital Object model (see @fils) for all vector files.
- [ ] Region
- [ ] Catchment
- [ ] Elevation
- [ ] Roadway
- [ ] Flow Line
- [ ] First Floor elevation
- [ ] Submitted model (inputs? outputs?)
- [ ] Synthetic power grid (As demo. replaceable by actual power grid)
- [ ] what else?
#### Model
[Where's the Flood?](https://docs.google.com/drawings/d/1qy4F-sDJV9n2Gcr8EKZwDhq3PtmYEDeXaIi4MtWeQJ4/edit?folder=1nThJj8Qw2QVnPf-eV2h-cu1bHqkO7M7-)
### Logical Model
Build a working ontology for the graph data (@torsten)
## Submission
## Crowdsourced data for the Network
GOAL: groundtruthing, prediction improvement, knowledge sharing
### Reporting App for Significant Flood Events
The Reporting App takes the geolocation of the reporting device and the date/time of the submitted event and contributes it into the knowledge graph.
Requirements:
- [ ] user profile - for managing reputation inside the social network. handling false positives and fake data? See Waze.
- [ ] profile types? Homeowner, Utility Manager, Civil Engineer, etc? as a proy for available event types
- [ ] geolocation data from the device
- [ ] date/time data from the device
- [ ] taxonomy of reportable events
- [ ]
#### What is the taxonomy for Event types? (Siddarth)
- [ ] Dam Breach
- [ ] Power Grid Failure
- [ ] Downed Power Line
- [ ] Flooded House
- [ ] ????
## Inferencing Tests - on any inntersection between a catchment and road.
- [ ] boolean flag for reclassifying flooded intersection
- [ ] model output value against the depth and flow rates of an intersection.
## Build for Model Submission (@siddarth - Jan. 2020)
### What parameters are required? units?

Computing file changes ...