Moderation rules can be used to automate the moderation process. These rules are created, edited, and deleted in the Moderate -> Moderation Rules section in the console. This is one of the most powerful features of Hyvor Talk.
You are advised to go through this documentation before using this feature.
How does it work?
Think of moderation rules as
a given value(Value),
an action(Action) will take place.
The type of the rule. Basically, defined what is matched to the condition. Hyvor Talk has 8 types, at the moment. Each type has one piece of data associated with it (Ex: User's name, comment, number of flags, etc.)
A value is given by moderators to match with the data of that type. If the match is successful, the given action will be executed.
All the numerical values should be equal to or greater than 0.
String values allow you to define multiple comma-separated values. Each value can contain any valid PCRE regex. However, if you use regex, limit it to basic regex, such as
[0-9], to avoid conflicts.
There are 4 actions for each type.
- Keep pending for moderation
- Mark as spam
See comment moderation to learn more about pending, spam, deleted, and approved comments.
Rules are executed on each comment based on the priority. High priority rules will be executed first. If a rule is matched and the action is called, the other rules are ignored.
Good to know:
- Most of the rules are executed when publishing or editing a comment. Some are executed on other events such as upvoting, flagging, etc.
- Moderation rules are executed only on unmoderated comments. For example, they are not executed on approved or deleted comments.
Words in the comment matches: Checks if words in the comment matches given comma-separated words. Executed when user publishers or edits a comment. Action is called if one or more words matches. This rule can be used to restrict some words.
Link domain matches: If any domain name of the links in the comment matches given values, the given action will be executed. This rule can be used to restrict several domains such as link shorteners to prevent social marketing through comments.
User's name matches: Matches the value with the user's full name (not username).
Number of links in the comment (more than): If the number of links in the comment is greater than the given number, the action will be executed. This rule can be used to prevent sharing many links in a comment.
Number of comment flags (more than): Logged in users can flag comments. If the number of flags per comment is greater than the given number, the given action will be executed. Setting this rule is recommended.
Number of downvotes (More than): Logged in users can downvote comments. If the number of downvotes of a comment is greater than the given number, the given action will be executed.
Number of upvotes (More than): Logged in users can upvote comments. If the number of upvotes of a comment is greater than the given number, the given action will be executed. This rule can be used to automatically approve comments that get many upvotes.
User's reputation (Less Than): All Hyvor users have a reputation in Hyvor Talk which is based on their contribution, upvotes, and downvotes. If the reputation is less than a given number, you can do something such as keeping comments pending for moderation. The reputation for new accounts is 0. Therefore, you can use this rule to manage comments by new users. The reputation of a guest commenter is also considered as 0.
If you set any of these rules, it is better to let your users (commenters) know about that to avoid frustrations. We suggest you add "Comments Note" in the Comments Note.
Here are some common use cases of moderation rules.
- To hold a comment for moderation if it contains one or more links
- Type: Number of links in the comment
- Value: 0
- Action: Keep pending for moderation
- To automatically delete a comment when it has more than 5 flags
- Type: Number of comment flags (more than)
- Value: 5
- Action: Delete
You can play with the other options and figure out the best ones for you. Currently, there's no way to see when and on what comments moderation rules are applied. That feature will be added in a later release if there's any public interest.