Adblock for FFXIV.
Certain symbols are turned into one space so the model sees multiple words instead of one. Previously "[RP]Hi" would turn into "RPHi" and be its own token. Now it turns into "RP" and "Hi", counting as two tokens. This change increased the model's accuracy. Also make "18", "http", "https", and LGBT-related words into stop words (meaning they're ignored). Each of these stop words made the model more accurate and reduced unwanted bias. Messages destined for ML are now normalised by the plugin in the same way the model's input is for training. This should make the results come closer to expected. |
||
---|---|---|
NoSoliciting | ||
NoSoliciting.Interface | ||
NoSoliciting.Internal.Interface | ||
NoSoliciting.MessageClassifier | ||
NoSoliciting.Tests | ||
NoSoliciting.Trainer | ||
.gitattributes | ||
.gitignore | ||
LICENCE | ||
NoSoliciting.sln | ||
README.md |
NoSoliciting
Adblock for FFXIV.
Summary
NoSoliciting filters chat messages and Party Finder listings based on
various built-in and custom filters. The built-in filters are updated
in the definitions.yaml
file and can be updated without a plugin
update. Custom filters are user-defined and are either
case-insensitive substrings or regular expressions.
All messages and listings filtered are logged in case of false positives, and there is a reporting mechanism in which users can report if NoSoliciting had a false positive or a false negative.