ReddiquetteAI Manual

Document Version 2

Frequently Asked Questions

What is ReddiquetteAI?
Like Karmangler, original-finder, and tweet_poster, ReddiquetteAI is a bot which makes automated messages on Reddit.

Why?
ReddiquetteAI was created to address violations of Reddiquette.

How?
ReddiquetteAI periodically scans the "Hot" and "New" queues of popular communities. It checks comment threads for terms and phrases which sometimes violate Reddiquette, such as "lol", "this", and "upvote". If it detects a term or phrase, it will calculate its confidence in its findings before making an automated message. Downvoted comments, short comments, and submission titles written in capital letters are more likely to receive a message from ReddiquetteAI.

What does reporting an error do?
If ReddiquetteAI mistakenly accuses someone of violating Reddiquette, writing an error report will help prevent similar mistakes in the future. When writing an error report, be sure to include a link to the erroneous message.

Does ReddiquetteAI ever upvote or downvote?
No. ReddiquetteAI never upvotes or downvotes other comments or submissions. See its "Liked" and "Disliked" pages.

Why didn't ReddiquetteAI respond to these comments?
ReddiquetteAI operates periodically. Also, it will only leave one message per comment thread. If it is prevented from commenting by Reddit's time limit, it will ignore that comment thread and move on.

Infrequently Asked Questions

What programming language is ReddiquetteAI written in?
ReddiquetteAI was originally written in Python 2.7.2. It will be rewritten in Python 3.2.2.

Will ReddiquetteAI's source code ever be made publicly available?
Maybe.