Gabor SzaboA while ago I volunteered to moderate one, later a few more tags here at DEV. For some reason the admins of the site also gave me some extra rights.
So I was blessed, or cursed, depending on the time of the day, with the ability to use some features on DEV to report spam.
I am not sure if this actually gives me more power than all the other users have. I think every user, yes, you too can report abuse by clicking on the 3 dots on the left-hand side of an article and then clicking on Report Abuse. It brings up a form where you have a selection and you can mark it as a spam. The only difference might be, that I see a bigger button to do so.
Anyway, in the last several weeks I started to post a lot of articles on DEV. With that I started to care more and started to watch the latest posts to report spam.
Observations
I started to see a pattern, that spam messages tend to be posted by new users. This might not be very surprising, after all a spammer probably registers an account and starts to spam immediately.
I assume it is rare that a user would post several good articles and then starts to spam.
The tool to help me
Using the API of DEV, or rather that of Forem, the engine used for DEV, I created a web site to help me locate spam messages. The site shows the most recently posted articles. Currently showing al the articles posted in the last 24 hours. It also highlights articles that have a higher chance of being spam.
The rules
I used several rules to highlight posts:
- An article by an author who registered less than 2 days ago.
- An article by an author who posted more than once in the last 24 hours.
- The first article of an author.
Obviously just because someone has signed up recently and already posted an article does not mean that s/he is spamming. Actually none of the above "proves" that someone is a bad actor or a post is a spam. So this application does not report anything to DEV. It only displays information to help a human decide which article is spam and which account to report.
Soon I realized that most of the articles that are the first articles of an older account are totally legitimate ones. Hence while meeting either of the first two conditions is marked by red tags on the article, meeting the last condition is marked by a green tag.
Most of the articles with a single green tag are genuine first posts of a person.
Reporting spam welcoming new authors
Every few hours, when I have some spare time I visit the page and start to scroll down focusing especially on articles that have a red or green mark. I pick the ones that look spam, check out the actual article and if necessary I report them.
I also pick the ones that have a green mark and try to comment on them. Sometimes just by congratulating them on their first article. Sometimes by suggesting improvements. e.g. recommending to add tags in general and language tags in particular, using Markdown better, or asking them to include links to what they refer to in their posts.
Welcoming new authors
Which brought me to the part where anyone could use this site. You too could check the site from time to time. Locate articles by first-time authors and welcome them!
People love it when you notice them.
Source
The source code is linked from the site.
