As some of you know, I've been working on a project that bears enough similarities to a dating site to be considered Generic Dating Site for the purposes of this post.
The field in which its denizens operate is one in which consent and trust is key. As such, one thing that I want to address is abusive people, and a way to handle them, while also accepting that it's feasible that malicious reports are more likely when they're done behind a keyboard.
The process that I have so far is this –
- Alice has had an abusive experience at the hands of Bob.
- Alice hits 'report abusive behaviour' on Bob's profile and fills in a few details.
- The system checks to see what relationship Alice has with Bob according to the site and allocates a points value to the report.
- Alice's account must be at least three months old to submit a report.
- If Alice and Bob are friends, 20 points are allocated.
- If Alice and Bob are friends-of-friends, 10 points are allocated.
- If Alice and Bob have no relationship, 5 points are allocated.
- The points value allocated is added to Bob's total.
- Once Bob's total reaches 50, a 3-month timer starts.
- Visitors to Bob's profile are shown a header saying that Bob has received reports of abusive behaviour, warning them to take care if they choose to meet Bob.
- If Bob does not reach 100 points within the three-month timer, Bob's total is reset to 0 and the warning is removed from his profile.
- If Bob reaches 100 points within the three-month timer, his account is suspended and site administrators are notified for further investigation.
- If the investigation shows that Bob has broken the terms of service in use of the site, an appeals process starts which may end with anything from removal of all points up to and including account termination and permaban.
- If the investigation shows that Bob has not broken the terms of service in use of the site (but understanding that it's possible that he may have engaged in abusive behaviour offline), his total is reset to 50, and the three-month timer continues and the process restarts from the stage at which a user attaining 50 points would reach.
The way I see it is that weighting reports in which the people involved have a social connection, we help to avoid malicious reports from people who've just had a disagreement in a comment thread somewhere, while also accepting that abusers are often known by their victims. We also try to remove any sense of 'scene celebrity' by working on a points basis.
I'm very interested to hear if anyone has any thoughts on this, other than OMG POOR MENZ. This is a first iteration after all, and weighting things in favour of people likely to be victims is a better start than weighting in favour of potential abusers.