issues
search
unsulliedcrowd
/
unsullied-api
0
stars
0
forks
source link
Basic functionality
#1
Open
steffansluis
opened
5 years ago
steffansluis
commented
5 years ago
Configuration:
[ ] Load an existing knowledge base schema.
[x] Define a example schema (ttl)
[x] Build schema logic (load schema)
Tasks:
[ ] Generate tasks to discover new entities and keep existing entities up to date.
[x] Build task logic (task generation, aggregation)
[x] Implement different types of tasks
[ ] Determine target personas for workers.
[ ] Determine relevant properties
[ ] Build persona logic (generate personas)
Crowd:
[x] Generate a custom worker interface to source workers through external means.
[x] Design a UI
[ ] Generate UI for question
[ ] Assign tasks to workers based on a configurable strategy.
[ ] Define strategy
[ ] Implement functionality to determine best option for assignments from a list of available workers and available tasks
[ ] Aggregate the worker contributions based on a configurable strategy.
[ ] Define aggregation strategy
Results:
[ ] Expose the generated data to the requester.
[ ] Store in CSV
[x] Reactively manage this process as the state of the knowledgebase changes.
[x] Make everything observable
[x] Show the requester information about the crowdsourcing process as it goes on
[x] Generate statistics from observables and display them om dashboard
steffansluis
commented
5 years ago
[ ] Finish up task logic
[ ] Use Redis to manage current db state
[ ] Generate Fix-Tasks based on current DB state
[ ] Generate Verify-Tasks based on current task state
[ ] Implement worker logic
[ ] Add preferredPersonas to config
[ ] Implement base crowd management strategy
[ ] Implement base task allocation strategy
[ ] Implement base task aggregation strategy
[ ] Expose results
[ ] Reuse current db-state from redis here
[ ] Build UI things
steffansluis
commented
5 years ago
UI:
[ ] Add configuration screen
[ ] Upload schema textbox
[ ] Implement task UI
API:
[ ] Implement reactive task generation
[ ] Implement random task allocation
[ ] Implement some kind of result aggregation
Configuration:
Tasks:
Crowd:
Results: