Dealing with evidence submission spam

[john-light]

In the current implementation of Aragon Court, anyone is able to submit evidence for a dispute during the Evidence Submission period. Currently the only cost to submitting evidence is the gas cost in ETH, which is a relatively low barrier unlikely to deter spam in cases where someone is highly motivated to interfere with the orderly adjudication of a dispute.

One idea I have to deal with this is to limit evidence submission to only the parties of a dispute (the entity who raised the dispute and the entity who performed the action that being disputed). If a third party wants to submit evidence for a case then they can try to convince one of the sides to the dispute to submit evidence on their behalf. Then if one of the sides is obviously guilty of submitting spammy/ low-quality evidence, the jury has a reason to rule in favor of the side with stronger evidence.

[marsrobertson]

Related: https://github.com/kleros/tokens-on-trial/issues/46

One of the ideas was an ability to mark evidence as "fake news".

But the attacker has advantage as they can submit any number of spammy content, so much effort to deal with all of it.

If parties are anonymous addressed on the blockchain, how to contact them "hey I have something for you"?

(Twitter, aragon.chat, Discord, Etherscan comes to mind)

[john-light]

If parties are anonymous addressed on the blockchain, how to contact them "hey I have something for you"?

Presumably there's enough context around the case that it would be possible to find contact info for involved parties. If not, posting the evidence in public somewhere and trying to get the attention of the people involved in the dispute may be the best that the submitter can do.

[Creativenauts]

I believe this is something that should be addressed. you could have good actors who choose to participate on a jury where evidence is submitted in the form of an executable which could potentially pose a huge risk and threat to jurors.

[john-light]

One potential way to deal with "evidence moderation" is to make this a paid service that jurors can opt-in to. Jurors could subscribe to an evidence moderation service that will review evidence and then hide "bad evidence" (according to some agreed-upon criteria) in the interface of jurors who subscribe to that particular content moderator. Jurors will then trust the evidence moderation service to 1) actually hide bad evidence so they don't waste time or expose themselves to threats by reviewing bad evidence and 2) not hide good evidence that jurors need to rule coherently.

[Creativenauts]

One potential way to deal with "evidence moderation" is to make this a paid service that jurors can opt-in to. Jurors could subscribe to an evidence moderation service that will review evidence and then hide "bad evidence" (according to some agreed-upon criteria) in the interface of jurors who subscribe to that particular content moderator. Jurors will then trust the evidence moderation service to 1) actually hide bad evidence so they don't waste time or expose themselves to threats by reviewing bad evidence and 2) not hide good evidence that jurors need to rule coherently.

Yea, that's not a bad idea at all, i actually really like that. Anyway we could create spam words as well so no one is able to use offensive language when submitting evidence?

[john-light]

Anyway we could create spam words as well so no one is able to use offensive language when submitting evidence?

I think jurors could self-select to filter out any kind of content they want. It also depends on what the evidence moderators are willing to filter for.