Hey, working on getting Osprey deployed for a platform but just have a query
One common way we might want to detect spam is by having something along the lines of:
messages = fetch_users_messages(limit=10)
similarity = compute_similarity(messages)
if similarity > 0.8:
ban_user()
However I'm not 100% sure what the best way to achieve this would be using Osprey.
This is primarily because I don't see a documented or implemented way for UDFs to query Druid for the message history?
Is the best solution just to put Redis/memcached in front of Osprey and query that, or is there some built in way to ask for the info from Druid itself?
Thanks!
Hey, working on getting Osprey deployed for a platform but just have a query
One common way we might want to detect spam is by having something along the lines of:
However I'm not 100% sure what the best way to achieve this would be using Osprey.
This is primarily because I don't see a documented or implemented way for UDFs to query Druid for the message history?
Is the best solution just to put Redis/memcached in front of Osprey and query that, or is there some built in way to ask for the info from Druid itself?
Thanks!