The API for spamicity.info uses a few functional endpoints.
Each endpoint takes a single Activity Streams activity as an
HTTP POST body. The Content-Type
for the POST body must
be application/json
.
The API methods require OAuth 1.0 for authorization.
You don't need to provide a token or get one; just get a consumer key and secret by logging in.
Ask if the activity is spam. The activity data is put through a Bayesian spam filter to determine its spamicity (thanks, Paul Graham!).
The result is a JSON document with a single object with the following attributes:
true
if
spam, false
if not. Based on the probability cutoff
configured at the server.Trains the activity as spam. Returns a training record, a single JSON object with the following attributes:
Note: After the initial training, it's more effective to
train a Bayesian filter only when its current results are wrong. So,
call /is-this-spam
first.
Trains the activity as spam. Returns a training record, a single JSON object with the following attributes:
Note: After the initial training, it's more effective to
train a Bayesian filter only when its current results are wrong. So,
call /is-this-spam
first.
Utility that shows how an activity will be tokenized. Possibly useful for debugging; also a nice idempotent method to test your OAuth keys with.
Returns a JSON document with an array of strings.