blocking people

[Ken-ichi]

I've fielded a few requests to block people in the last couple months,
so I guess it's finally time to build blocking functionality, an
unfortunate but necessary component of any sizable online social
network. Before we get building, I'd like some feedback from you all
on how this should and shouldn't work.

Here's what I'm thinking. Blocking will:

* prevent the blocked user from commenting on your observations,
journal posts, or any content you author
* prevent the blocked user from identifying your observations
* prevent the blocked user from messaging you
* prevent you from receiving a notification if the blocked user @ mentions you
* stop and prevent the blocked user from following you

Blocking will NOT:

* prevent the blocked user from viewing your profile, observations, or
any other content you author
* prevent you from seeing the blocked user's content in public places
like http://www.inaturalist.org/observations

Basically, I'm taking my cues from Twitter, where everything you do is
publicly visible by default, just like on iNat. We want you to be able
to ignore people you don't like, but iNat is public by default, so
there's no reason for us to prevent someone from doing something they
could just as easily do by signing out and viewing the site
anonymously. There are, of course, more serious personal security
considerations on iNat since everyone is constantly sharing places
they've been, but that's a risk every iNat user takes and should
understand.

Anyway, if you have any thoughts on how this should work or other
things blocking should or shouldn't do, please let me know. If you'd
prefer not to talk about it in public, please email us at
he...@inaturalist.org.

-ken-ichi

[Donald Hobern]

Hi Ken-ichi.

In general, your approach sounds fine (using Twitter as a model).  The one part that made me unsure was "prevent the blocked user from identifying your observations".  I realise that this may be precisely the driver for some people wanting this feature, but it seems a slippery slope.  One of the real strengths of iNaturalist is that users can identify any other observation.  This ensures that the number of observations is maximised.  A model which moved towards users deciding who can and can't identify their records (and where potential identifiers may be frustrated by finding that a significant number of records block their attempts at identification) could destabilise the model.  I realise there is a significant distance between users blocking one or two other users and users only admitting one or two users to identify their records, but this is a significant change in philosophy.  If nothing else, it opens a door for someone to create and verify garbage identifications, while rejecting any disagreement.

I can see the value of allowing users stop having to see comments from users they do not wish to have commenting on their observations, but could this just be by hiding some comments from the view when the blocking user is logged in?  Does it really need to lead to an increase in variation on what users can do with different observations?

It might help to have a better understanding of the type of situation which is leading to this request.  Since users can already stick with their own identifications rather than community identifications, one major sensitivity has been addressed.  Most other issues I can think of that would merit a change like this seem rather to be the kind of reason for considering deleting offending users.

Best wishes,

Donald

--
You received this message because you are subscribed to the Google Groups "iNaturalist" group.
To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist...@googlegroups.com.
To post to this group, send email to inatu...@googlegroups.com.
Visit this group at https://groups.google.com/group/inaturalist.
For more options, visit https://groups.google.com/d/optout.

[Tony Wills]

Hi all.

I generally agree with Donald's points, it would be useful if we had some actual cases (with names omitted ;-) that sparked the requests to Ken-ichi.  I have dealt with a very large number of observation flags and certainly do not see any pattern of abuse or harassment being reported there - but of course if it is a personal issue, they may not want to use the public observation flag system?  Usually it is just newbies trying out the system who enter daft identifications, and nobody needs to be blocked.

I have absolutely no problem with creating ways for people to ignore the messages from others, such as sending pings and personal messages to the bit bucket

But I think we need to decide whether this is a citizen science project or just another social media site for our entertainment?  If it is just to be just a social media site, with the hook being the environment, then it doesn't matter so much - a system of friends and cliques etc is not a problem.  If an essential part of the site is the data, then that should be open to be contested with both identifications and comments.  If we restrict identifications and comments on observations, I don't think those obsevations can reasonably be considered 'research grade' and shared with other data aggregators.

The system of flagging comments should be enough to deal with personal comments and highlight patterns of abuse.  Can observers delete comments, or only curators?

If someone is being deliberately disruptive, then no doubt they should be site banned (after a warning perhaps).  Though you then just get into a war of them changing usercode/email-address/ip-address to get around the block - you might just create ongoing work for admins.

If people don't get along, or have no respect for each other's work, blocking each other would seem to be counter productive - science lives by testing and re-testing theories/arguments and scrutinizing data, not by ignoring-things/banning-people we disagree with.

If the site is willing to host observations and projects whose observations are locked, or can only be identified by project members or other selected individuals, I have no problem with that - again I feel we shouldn't publish those observation identifications as 'research grade'  - the data obtained might even be better than crowd sourced identification, but it needs to be flagged as uncontested or identified by expert group xyz etc.

Again some case studies might help me understand the problem we're looking to solve.

Tony.

[Charlie Hohn]

Tony, I get what you are saying and you make some valid points. However, my concern is that there have been some very nasty streaks of harassment on the Internet as of late, including cases where the harassment might not be obvious to admins/curators on this site because it's spilling over from another site or real life. I don't think iNaturalist should be a place where we allow that sort of thing to happen, and while we have a great community, it might not always be obvious to us if someone is being harassed by a stalker, abusive ex,bully, etc. I believe the iNat community is large enough that we can  absorb a minor amount of blocking without it affecting the integrity of the ID system at all. I guess I can imagine a situation where someone blocks everyone that disagrees with them, but if that were happening imho it would be really obvious and we could deal with it. If it really seems like a problem there could be an internal flag so if someone blocks more than say 20 people the situation gets reviewed.

So yeah I think what Ken-Ichi is proposing works. It's a shame it has to come to this. I haven't observed any harassment at all, some heated discussions that become cranky I suppose, but it's been pretty limited.

--

============================
Charlie Hohn
Montpelier, Vermont

[Charlie Hohn]

also note that Ken-Ichi's system does not stop someone from flagging the observation of someone who banned them. If someone bans me for disagreeing with their ID and has their friends agree in an abuse of the system, I would just flag the observation.

[Donald Hobern]

Hi Charlie,

I agree with the points you've made, but would like some better idea of the kind of issues involved.  

If the issue is around stalkers, abusive exes, bullies, etc., then I would support a very aggressive approach to dealing with it (and I suspect that the kind of blocking proposed may be insufficient).  

If it is about people not liking the judgment of other observers on their observations, then my feeling is that we should still consider and investigate what is going on and how it can be managed, but blocking still doesn't necessarily seem the right approach to me.  

I'd like to understand what situations are envisaged which sit between these two cases - it may be that Ken-ichi's proposal perfectly hits the sweet spot, but I can't tell right now.

Donald

[Charlie Hohn]

I am curious as well, since I haven't observed the abuse. (that isn't saying it isn't there! I just haven't happened to find it)

[AfriBats]

Hi everyone

Just a quick one to say that I share Donald's and Tony's concerns, and that I generally would not like to see users given the power to exclude other users from IDing, or publicly commenting on, their observations for said reasons.

I'd be fine if users would have the option to block personal messages, and making other users "invisible" to them.

Best, Jakob

[Ken-ichi]

Hey folks,

Thanks for the feedback. Obviously I won't go into specific cases, but
in general terms, the primary kind of situation I want to deal with is
what Charlie described: preventing harassment that falls short of
behavior prohibited by our Terms of Service, but is still undesirable,
e.g. superficially innocuous communication from someone you've had
less innocuous interactions with in the past like an ex or bully. I
don't think this behavior is becoming more common, it's just becoming
more probable the larger our community gets. I think for most people
who have experienced harassment like this in the past, *any*
intentional communication from that person would be considered harmful
to the point of not wanting to use our site, and in my mind, that
includes identifications, even if there are no notifications.

Another less alarming use case: let's say there's a user who sends you
a message *every day* asking for help with a new insect he saw. You've
asked him to stop, or at least ease off to once a week or something,
but he won't. There's probably nothing malign about this user's
intents, but you and he have different understandings of what
constitutes an appropriate level of communication between strangers.
The constant updates are so annoying that it makes you not want to use
the site.

I personally think the fears some of you have about people blocking
others just because they don't trust their identifications are
unwarranted. I suspect that would be rare, and in theory its effect
becomes vanishingly minor the more identifiers there are. Plus,
similar manipulations of our crowdsourced identifications are already
possible: anyone can just make a bunch of extra user accounts and
agree with their own identifications. To my knowledge, that's happened
exactly once in the history of iNat, and we took action to stop it.
The incentives for manipulating the system are just so low that I
doubt that kind of behavior will ever be a serious threat.

We could also just make a policy that states that blocking is not
intended as a way to opt out of community IDs, and if we suspect
someone of doing so, we'll investigate and possibly suspend their
account.

-ken-ichi



On Tue, Apr 19, 2016 at 7:42 AM, Charlie Hohn

[Donald Hobern]

Thanks, Ken-ichi - and it's easy for me to sit here on the outside and be dogmatic.  I'd suggest that one possible way to proceed would be to have a "Mute" option which allows a user from that time forward not to see updates from a user they wish to avoid (comments hidden, no mail delivered) but leave it possible still for anyone to comment and identify on any observation.  It seems a safe halfway point and would still allow the blocking of identifications if problems remained.  

Best wishes, Donald

[Lynn Watson]

How about a warning system first. A user may not realize they are being offensive in some cases. 

[AfriBats]

On Wednesday, 20 April 2016 08:56:27 UTC+2, Donald Hobern wrote:

Thanks, Ken-ichi - and it's easy for me to sit here on the outside and be dogmatic.  I'd suggest that one possible way to proceed would be to have a "Mute" option which allows a user from that time forward not to see updates from a user they wish to avoid (comments hidden, no mail delivered) but leave it possible still for anyone to comment and identify on any observation.  It seems a safe halfway point and would still allow the blocking of identifications if problems remained. 

That's quite similar to what I suggested above and I'd vote for that.

Best, Jakob

[Charlie Hohn]

i vote that we do what Ken-Ichi suggested but watch for abuse, and if abuse of the ban system to manipulate IDs happens THEN we reevaluate. I don't think it's going to happen.At least not enough to matter or justify the risk that we allow people to be driven off the site by harassers. 

If you don't believe me google the words "gate harassment" . There are some nasty people out there. Better safe than sorry IMHO. Glad they haven't been to the site but I've been watching the stats and inat is hitting a growth spurt. Great news but we need to do a good job managing it.

--

You received this message because you are subscribed to the Google Groups "iNaturalist" group.
To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist...@googlegroups.com.
To post to this group, send email to inatu...@googlegroups.com.
Visit this group at https://groups.google.com/group/inaturalist.
For more options, visit https://groups.google.com/d/optout.

[jesse rorabaugh]

My suggestion is the following:

Don't inform the person that they were blocked, the system should just act like the person is ignoring you. Doing otherwise will just encourage making a dummy account, or otherwise finding ways to cause trouble. Also, with some of the odd messages I have got from new users, such as the guy who sent me messages unhappy that I was agreeing with his identifications, I fully expect that a large percentage of heavy users will eventually get blocked by someone. Telling them they were blocked is likely to discourage them which is generally something the site should avoid doing.

When the blocked user sends messages, the recipient never gets them. The sender just is left assuming that they got through but were ignored. 

Don't block identifications, but don't display any comments which come along with the identifications. Let them type comments in, maybe even show the comments to the person who submitted them so they think the system is working as usual, but don't display them to any other users. There isn't much you can do to harass a user through identifications, and it is a very public forum so I highly doubt continuing to allow identifications will be a problem. 

[Jon Sullivan]

I've been thinking about this thread for a while. It's not an easy compromise to find. I'd hope that whichever tool is implemented is hardly ever used.

Ken-ichi's solution seems simple and adequate. The obviously inappropriate comments can still get flagged by everyone as usual. If someone starts slamming me with lots of inappropriate comments, or makes even one comment that is really offensive, it makes sense for me to be able to turn on a block.

I've been thinking about Jesse's suggestion to make it unclear to the abusive user that they've been blocked. It sounds tricky to implement given that we'd need to maintain the equivalent of private comments that only the person blocked could see, to hide from them the fact that they'd been blocked. I also wonder if this would be counter-productive in most cases.

What would I see if someone had blocked me and I tried to add an identification or comment to one of their observations? Would I see an explicit "You can't do this because the user blocked you" message, or some more vague "Sorry but you cannot add comments or identifications to this observation at this time"? Or would there simply be no comment box or "Agree" link available to me?

In most cases I expect it's useful if it's clear to the blocked user that they've been blocked. If two people have such a broken relationship that one person will make new accounts to try and get around a block, then a public online forum like iNat is probably not the place for either of them. I'd hope that much more often someone will have said inappropriate things that have offended someone else, perhaps even inadvertently, and they get blocked by that person. In that case they should know about it and will hopefully shrug it off and learn to be more polite next time. In this case, I agree with Lynn that it would be handy to have a warning first. If I flag a comment on one of my observations as offensive, it should remain easier for me to flag that comment and trigger a warning to the person who made it, than jumping straight to an all-out block.

We did have one case in New Zealand several years ago of one user being so offended by one comment that she deleted her account outright (with lots of observations all IDed and commented on by the NZ community). The comment was nothing malicious, just an off-colour attempt at humour that was interpreted too literally. She thought about it and eventually returned to the site (and I think we managed to restore her observations from a backup). A block tool would have been helpful in this case, and I don't think there would have been any harm in iNat alerting the commenter that they'd been blocked.

Rather than Jesse's ignore approach, I wonder if the blocker should even be given the option of personalising the message to the user being blocked. If someone blocks me, I'd like to know why. And if I blocked someone, I'd want to tell them why. I could always message them one last time then block them.

Cheers,

Jon

[Charlie Hohn]

" If two people have such a broken relationship that one person will make new accounts to try and get around a block, then a public online forum like iNat is probably not the place for either of them. "

disagree. Sometimes this is true but there are those nasty cases where one person is being vile and the other person really is blameless.  Again, referring to more serious things like abusive relationship partners or stalkers. It's rare, but it does happen, I know people who have been targeted by it and it isn't fair to punish them for the actions of their abuser.

Sorry to take such a dark tone. hopefully it simply never comes up

--

You received this message because you are subscribed to the Google Groups "iNaturalist" group.
To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist...@googlegroups.com.
To post to this group, send email to inatu...@googlegroups.com.
Visit this group at https://groups.google.com/group/inaturalist.
For more options, visit https://groups.google.com/d/optout.

[Tony Wills]

A simple blocking mechanism can not possibly be effective in stopping subtle harassment by someone intent on doing it, as it is trivial for anyone to create a new account (even if they need a new email and IP address).  There is only one way to stop subtle harassment - be invisible to the person harassing you, so they don't know who to harass (ie create an anonymous account).

The main purpose for creating a blocking mechanism is so that site admins and curators don't have to do the impossible task of deciding whether someone is subtly harassing someone else.  Give them a block button, so that they can block whoever they feel threatened by.  But this is also inherently un-policable, we can not know whether their concerns are well founded, almost by definition - ie if it is obviously harassment we would simply ban that usercode anyway - this button is for other cases.  The subtle harassment is from the fact you know the other person is there, watching you, following you.  They don't have to do anything, even adding legitimate/useful identifications can be part of it - they're just letting you know they are there.

But what we do know is that such a button will be used casually to block people who are annoying in any way (from disputing ones expert opinion, to blocking someone who is trying out the site and adds something dumb).  I see a lot of flags of this nature, often anything unwanted will be marked as "spam" which has the satisfying effect of removing the unwanted item and blocking the other user.  So what are we going to do? Overturn blocks we think are unwarranted?  Warn the blocker that they are using the tool excessively because we don't think this is real harassment (how do we know?) ?

We can not solve a "subtle harassment" problem.   We can give the user a small amount of help by allowing them to block pings and personal mail (just discarding them as though they are being ignored) - I think this feature can be used without restriction, if you want to ignore people, do it to your hearts content.
And people will also use it because someone else simply annoys them, a pity, but not a great problem in general.

It would be nice if the user could also block seeing comments by others on their observations, but this would perhaps be more painful to implement (and create a few confusing conversations when others respond to that 'invisible' comment).  Ignoring identifications would be problematic (the observation suddenly becomes research grade with no other apparent id, or more likely goes from being research grade to casual with no apparent conflicting id?).  If the blocking is to be overt - ie the person being blocked will know about it, then it is of no real use - if that person really is intent on harassment. they will simply try other things (ie a new account).

To sum up my view:
Overt harassment (deliberate abuse or disruption through pings,mail,comments,ids) - ban the user, repeat for each new harassing usercode.
Subtle harassment, we can not solve almost by definition.  Only real solution is to become invisible to the harasser.
A block button to discard pings and mail will be much used (and very satisfying for some :-), but can't actually stop subtle harassment.
Blocking comments and identifications probably can't be done in a way that doesn't alert the person being blocked, so would be ineffective in blocking the subtle harasser - they would just create another account.

-- Tony.

[Sheri L. Williamson]

Has there been any progress on a block/mute option? I'm being "stalked" by a user and would really like to prevent him from following me, getting notifications about my sightings and identifications, and adding IDs to my sightings.

[Tony Iwane]

Hey Sheri, can you email us at he...@inaturalist.org with some more details about what's happening? We can look into the situation for you.

Tony

[James Bailey]

Blocking users from identifying your observations sounds like a bad idea for many reasons.

If an ID is spam, it can be reported. Otherwise such IDs should be taken as meaningful content. I don't want stubborn users blocking input because they can't take comments or suggestions. That isn't how iNaturalist works.

[Frank Fogarty]

I've just realized that I have been blocked by a user when I went to click 'agree' on an identification and it bounced. This seems to be a case of the user not liking their identifications corrected, as I have never exchanged a message with this user, rudely commented, or otherwise interacted with them in any way other than correcting IDs and typically providing field marks for those corrections. I did get one snarky response comment to a correction a while back that more or less said "I don't care if the ID is wrong, this is the only picture I have labeled as that species so I am keeping it". 

This seems to illustrate the problem of allowing the block function to prevent users from contributing to the identification process. Users have the ability to block experts or anyone that disagrees with their identification and are able to get observations to Research Grade without full community vetting. To that point, a few minutes after that I stumbled on a misidentified observation that is Research Grade from that user and there is seemingly nothing I can do about it.

On Monday, April 18, 2016 at 6:31:20 PM UTC-7, Ken-ichi Ueda wrote:

[Ben Phalan]

This seems likely to be rare enough that posting a link to these observations here in the Google Group would be enough to get a few other people to look at it and correct the ID.

[Tony Iwane]

We specifically say that blocking should not be used to stop someone from IDing your observations, so if this is the case it would be a misuse of Blocking. Frank, if you want to email us at he...@inaturalist.org with some more details, I would like to look into it.

That being said, the person clearly doesn't want to hear from you, so putting the URL(s) of their observations on the Google Group is probably not the best response and would likely exacerbate the problem, I think.

Tony

[Charlie Hohn]

if someone has blocked you can you still @ other people in their observation? Depending on the situation it might help get more IDs or might ust piss the person off more. If the person isn't interested in community ID they should turn it off, not harass people who disagree with them. Though if they are knowingly wrong and putting in wrong data to inflate their species count they are being absurd (or trying to cheat for a class) and iNat probably isn't a good place for them.

--

You received this message because you are subscribed to the Google Groups "iNaturalist" group.

To unsubscribe from this group and stop receiving emails from it, send an email to inaturalist+unsubscribe@googlegroups.com.

To post to this group, send email to inatu...@googlegroups.com.
Visit this group at https://groups.google.com/group/inaturalist.
For more options, visit https://groups.google.com/d/optout.

[Ian Toal]

This is all kind of depressing, but as Ken-ichi Ueda says, as the number of users grows, the chances of nastiness also grows. My only suggestion is rather than trying to design a system that covers every possible contingency, start simply. If it works, good. If it doesn't then modify it. The Twitter option sounds like a good place to start. 

[Tony Iwane]

Hi Paloma,

In our Community Guidelines we explicitly say "Blocking is not a way to stop identifications from people you don't like or trust. You can address problems like that by opting out of the Community IDs...If we (the site admins) find that you are using blocking to silence identifiers you don’t trust, we will investigate and we may suspend you."

If you can send the URL of the observation in question to he...@inaturalist.org, we can look into it.

Tony Iwane

On Sat, Nov 17, 2018 at 4:36 PM paloma <margaret...@gmail.com> wrote:

Today I saw an observation where a very helpful identifier was subjected to rude comments. The identification was not off-the-wall, as in calling a mammal an insect or something, but a disagreement as to whether something can be identified to subspecies or not. I found out afterwards that the observer has since blocked that identifier. I have no involvement in the particular observation, but I was kind of shocked at the comments, just because I expect people to have a reasonable discussion. Since I am not privy to other communications between the observer and identifier, I am not going to flag it. But as a user concerned that useful knowledge may be made unavailable to the community by inappropriate blocking where opting out of community ID would suffice, I feel cheated.

--
You received this message because you are subscribed to a topic in the Google Groups "iNaturalist" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/inaturalist/KsTKdb7wYS0/unsubscribe.
To unsubscribe from this group and all its topics, send an email to inaturalist...@googlegroups.com.

[Patrick Alexander]

I guess I'm still hoping for a "don't show this user's observations to me in explore / identify type of function... which is not to say that the kind of blocking proposed here is not necessary in some cases, but that there's an additional kind of blocking that would be helpful in a different set of cases.

[Mark Tutty]

You can filter to only show that users observations, and then quickly go through and mark them all “reviewed”. Then take that part of the filter off again, and review the rest as normal.

 

cheers
Mark Tutty
kiwif...@gmail.com

--

[Mark Tutty]

I think we need an additional “mark all as reviewed” on the identify page... one that works for ALL observations based on the current filter (ie “mark ALL from ALL pages as reviewed”). It would likely need a confirmation in case it is selected by mistake, and it probably should NOT have a toggle effect, ie there should not be a corresponding “mark ALL from ALL pages as un-reviewed”.

 

Actually... maybe make it for just 10 pages, as there could be consequences from over-reaching back too far! Or alternatively, only have the option available if the number of pages based on the filter settings is less than a set amount, say 10... that way it is not present when looking normally at the identify pages, but when looking at a filtered version of it, the ability to bulk mark obs as reviewed or un-reviewed comes into play...

 

Hmmm not as simple as I first imagined...

 

cheers
Mark Tutty
kiwif...@gmail.com

 

From: Patrick Alexander
Sent: Wednesday, 21 November 2018 11:06 AM
To: iNaturalist
Subject: [inaturalist] Re: blocking people

 

I guess I'm still hoping for a "don't show this user's observations to me in explore / identify type of function... which is not to say that the kind of blocking proposed here is not necessary in some cases, but that there's an additional kind of blocking that would be helpful in a different set of cases.

--

[Patrick Alexander]

For what it's worth, I think a setting like "don't show observations from user [...]"--and probably also "don't show observations from project [...]"--might also solve a lot of the issues with student projects and AI suggestions that come up in discussions here. I think people can get a little fixated on the idea that there are errors in the data--and we must correct them! Well, in terms of my own user experience it's really only a problem to the extent that I'm interacting with that bad data. I don't really mind that the bad data exists. I mind if I end up getting suckered into data cleanup tasks that I don't find rewarding or interesting, and the amount of time I spend on iNaturalist doing data cleanup is going to be proportional to how rewarding / interesting I find it. For instance, I cut down on a lot of the class project issue for my own use by creating a user-defined place that excludes El Paso. I'm sure the classes are still generating big piles of bad data for El Paso, but I don't feel a responsibility to fix that problem and I know I'll get irritable and frustrated if I try to fix it. If I can filter it out, that's good enough for me.

Right now we don't have a lot of great tools for that kind of filtering. I agree that being able to mark larger sets of observations as reviewed would help. I use "mark all as reviewed" a fair amount, but it gets really tedious paging through a big pile.

[paloma]

Patrick, might setting up a project with your geographical interest and accessing the Identify tool from within the project work? Projects already have a function to exclude users.