antispam:plugin:neuralnetwork

Log real-time per-rule statistics for each email, allowing admins to easily analyze rule performance and identify false positives/negatives.

warden --task=antispam:plugin:neuralnetwork
Option
 Value Default Description
--neuralnetwork_data_dir <string> /etc/mail/spamassassin/.neuralnetwork Where the NeuralNetwork plugin will store its data.
--neuralnetwork_min_text_len <digit> 256 The minimum number of characters of visible text required to run prediction or learning on a message.
--neuralnetwork_min_word_len <digit> 4 The minimum token length considered when building the vocabulary and feature vectors.
--neuralnetwork_max_word_len <digit> 24 The maximum token length considered when building the vocabulary and feature vectors.
--neuralnetwork_vocab_cap <digit> 10000 The maximum number of vocabulary terms to retain; least-frequent terms are pruned when exceeded.
--neuralnetwork_min_vocab_hits <digit> 10 The minimum number of tokens in the email that must exist in the vocabulary for prediction to run.
--neuralnetwork_cache_ttl <digit> 300 The time-to-live in seconds for the in-memory vocabulary and model caches Set to 0 to disable caching.
--neuralnetwork_min_spam_count <digit> 100 The minimum number of spam messages in the vocabulary required to enable prediction.
--neuralnetwork_min_ham_count <digit> 100 The minimum number of ham messages in the vocabulary required to enable prediction.
--neuralnetwork_spam_threshold <int> 0.6 Prediction values above this threshold are considered spam.
--neuralnetwork_ham_threshold <int> 0.4 Prediction values below this threshold are considered ham.
--neuralnetwork_learning_rate <int> 0.1 The learning rate used by the underlying FANN network during incremental training.
--neuralnetwork_momentum <int> 0.1 The momentum used for training updates.
--neuralnetwork_train_epochs <int> 50 The number of training epochs to perform when learning a single message.
--neuralnetwork_train_algorithm <string> FANN_TRAIN_RPROP The algorithm used by Fann neural network used when training, might increase speed depending on the data volume.
--neuralnetwork_lock_timeout <digit> 10 The maximum number of seconds to wait for the exclusive training lock before giving up and skipping the learn operation. Set to 0 to wait indefinitely.
--neuralnetwork_stopwords <string> the and for with that this from there their have be not but you your A space-separated list of stopwords to ignore when tokenizing text.
--neuralnetwork_autolearn <1|0> 0 When SpamAssassin declares a message spam or ham during the message scan, and launches the auto-learn process, the message is autolearned as spam/ham in the same way as during the manual learning.
--default <yes> Reset all settings to their default values.
--default_option <option> Reset a specific setting to its default value.
--reload <yes> Reload the service after saving settings.

Examples

// set the neuralnetwork_autolearn
warden --task=antispam:plugin:neuralnetwork --neuralnetwork_autolearn=1 --reload=yes

// reset neuralnetwork_autolearn to its default value
warden --task=antispam:plugin:neuralnetwork --default_option=neuralnetwork_autolearn --reload=yes

// reset all settings to their default values
warden --task=antispam:plugin:neuralnetwork --default=yes --reload=yes