Most likely not a Lens - Discussion on Public/Private collections

by Alpha_Aurigae

The arc looks like a lens in infrared but it is probably a starforming region

Posted November 7, 2014 10:02 PM
by Capella05 moderator

'Lenses' are my thing 😃

No sign of a counter image, and there are already indications of star formation within the galaxy.

I will put the 'blueness'down to star formation regions. There is also a foreground star just to the right of the bright region on the left. The galaxy is also one we would not necessarily associated with gravitational lensing. More inclined to think collisional ring.

No lensing for me 😃

Posted November 7, 2014 11:09 PM
by Capella05 moderator

On another note - we have not classified UKIDDS images for about 6 months - how did you come across this one?

Posted November 7, 2014 11:16 PM
by Alpha_Aurigae

Hi Julianne,

I am checking dozens and dozens of collections in order to tag NGCs, UGCs, MRKs. In the meantime I am also looking for interesting images. I found this one in a collection
(but whose I don't know anymore). There are a number of people who have got this galaxy in their collections.

Posted November 8, 2014 6:04 AM
by Capella05 moderator in response to Alpha Aurigae's comment.

The best way to find an interesting object is to classify!

As I have mentioned before, user collections can be very personal and some volunteers may not feel comfortable with someone else looking through them.

Also, it is not necessary to find all the NGC, UGC objects - there is already a catalogue for them 😉 Just tag them as you classify them 😃

Posted November 8, 2014 12:09 PM
by Alpha_Aurigae

I beg to disagree on this case. Volunteers' collections can provide a wealth of information. I found that out on Planet Hunters. I discovered a number of RR Lyrae stars (Type AB and C) that way. They will be included in a new scientific paper Robert Szabo is working on. I also found a few cataclysmic variables, one of which is described here: http://arxiv.org/abs/1308.0369

In most cases the volunteers collect the best of images and if you want to appreciate the beauty of the Universe you can just have a look at them.

Now if there is a need to keep certain collections private, shouldn't there be a private button? Maybe something for DZM to think about. Just like on Facebook, you decide whether or not other people get to see your collections.

Posted November 8, 2014 12:53 PM
by vrooje admin, scientist

First off: everything below is my personal opinion and not an official position.

I have mixed feelings about the notion of private collections. On the one hand, I definitely understand that feeling of not wanting to share a work in progress before it's ready. On the other hand, keeping things in a drawer basically makes it impossible to do collaborative science.

Also, I think there are two uses here: the first is appropriating a collection in its entirety; the second is identifying an object that is part of a collection but that is also interesting on its own. To me they're different.

If I came across a collection of objects that together represented something that I found scientifically interesting, and I wanted to investigate that science further, I'd contact the collection owner to see what they were already doing with it. Once we were talking about the science we'd go from there on whether and how to collaborate. But here as in any other academic collaboration I'd see the idea behind the collection as the collection owner's, and if it's an original idea they should get the credit for it (even if I was never able to contact them and went ahead anyway). People who intentionally steal other scientists' ideas should be named and shamed, and one advantage of openness is that it's easier to see whose idea something originally was.

On the other hand, a collection isn't the same as a Facebook photo album where someone is uploading their own photos: no user owns the objects in a Galaxy Zoo collection. If I come across a new voorwerpje because it's on the Talk homepage as part of a collection entitled "Asteroids", then I have no problem with starting a discussion about the new voorwerpje, adding it to my own collection of voorwerpjes, etc., whether or not I'm able to contact the owner of the Asteroid collection.

No matter what the decision is on the principle, I suspect the ability to make collections private is a significant enough development effort that it won't happen in this version of Talk. So I think openness is the default for now and we all just have to add ethical behavior on top of that. 😃

Cheers,

-Brooke

Posted November 8, 2014 2:55 PM
by Capella05 moderator in response to vrooje's comment.

Thank you Brooke for your opinion 😃

I think it is also important to point out that there are two different type of collections - a user defined one when the volunteer has to manually add an image to a collection they have created, and the automatically generated collections geared by hashtags.

For me, the latter suggests group ownership - the former 'belongs' to the volunteer who created it. Perhaps I have been studying too much philosophy lately, but surely the act of a volunteer taking the decision to create and add a image to a personal collection, does denote a somewhat benign form of 'ownership' within that collection? Yes, the image belongs to Galaxy Zoo - but as you pointed out we have no way of knowing the purpose that the collection was created for, unless it is stated.

It would be nice to make more of a differentiation between the two in the next iteration of Talk.

I will be honest - As a 'classifier' the one thing I do feel uncomfortable about, is the thought of someone else skipping the 'grunt' work of classifying, just to 'appropriate' the images I have deemed worthy for a collection.

The above is my personal opinion 😃

Posted November 8, 2014 3:30 PM
by JeanTate in response to Capella05's comment.

My 2 cents: scouring Collections for objects similar to a class of (relatively rare) objects you're doing research on is quite effective, in finding candidates. Over in RGZ, this is how I began my search for SDRAGN candidates*. Many objects can be found in more than one Collection, created by more than one zooite. My understanding is that Collections was developed, as a capability, to permit the kind of thing I just described; individual citizen scientists will notice patterns and Collect accordingly; professional scientists can mine these Collections for patterns that may have escaped their notice, and be able to characterize the patterns in terms of theories of galaxy formation (say). It's a way of realizing 'wisdom of crowds', similar to "What's this?" but extended to multiple objects (each one of which may be quite unremarkable).

Thought: has anyone looked at the overlap between Comments/Discussions and Collections? For example, in a not-too-small Collection, how many objects were Commented on? mentioned in Discussion threads? One feature of Talk that hinders such an investigation is the lack of a timestamp on when an object was added to a Collection (Comments and posts in Discussions have timestamps).

*later this became pretty much unnecessary (long story)

Posted November 8, 2014 4:22 PM
by Capella05 moderator in response to JeanTate's comment.

It might be effective, but not necessarily ethical.

How do you know the collection you are raiding is not data for a students term paper? Or a fellow CS long term project (for example the 'Irregulars' project)?

Posted November 8, 2014 4:45 PM
by williamaskew in response to Alpha Aurigae's comment.

I have to stand with AA on this ive a thousand hubble deep oranges you can go through if you want AA. I havent checked them for a couple of years but knock yourself out. I think a fresh set of eyes always helps.
Billy

Posted November 8, 2014 5:05 PM
by Capella05 moderator in response to wtaskew's comment.

That is the thing! @wtaskew - you have just given permission for Abe to look through your collection - that's cool 😃 - but that is somewhat different from looking through other volunteers collections, and posting the objects you have found without the owners knowledge.

Thinking about it, an acknowledgement to the collection owner could be nice 😃

once again my personal opinion

Posted November 8, 2014 5:25 PM
by Budgieye moderator

I thought that collections in Talk was the equivalent of posting to a thread in the Forum, but we could post to several collections at the same time.

I try to acknowledge if someone has a Good Idea. We can always look at the comments about object to see who was first to have the the Good Idea.

I think we should share, like SDSS has done with us, and Galaxy Zoo has done with the world.

Posted November 8, 2014 6:16 PM
by Capella05 moderator in response to Budgieye's comment.

I totally agree with sharing 😃

I personally have no 'user' collections (apart from the ones I have created for tutorial purposes), as I feel what we do is all part of a group effort 😃

I was just trying to stand up for the volunteers who thought otherwise, and were trying to conduct independent research.

Posted November 8, 2014 6:31 PM
by JeanTate in response to Capella05's comment.

It might be effective, but not necessarily ethical.

Something we should discuss, I think.

How do you know the collection you are raiding is not data for a students term paper? Or a fellow CS long term project (for example the 'Irregulars' project)?

I don't, and can't (unless they say so in their Collection write-up). What's in Talk - Comments, Discussion threads, Collections - is in the public domain. As far as I know, there are no limitations on access to it, and in one sense zooites are more restricted than those who have not registered (non-zooites did/do not have to agree to anything, re use of anything they find here).

In any case, Christian+ (2012) seem to be clear that professional astronomers can - and will - 'raid' Collections (or anything else they find in Talk) at will; for example, a Collection (or set of Collections) may be used as a machine learning training set without any notification or communication at all*.

Further, professional astronomers - whether on a Science Team or not - have the capabilities to 'raid' Collections far more quickly and effectively than any of us ordinary zooites ever will. And, unlike catalogs (etc) of Zooniverse data (such as Galaxy Zoo 1 data release - "Anyone making use of the data should cite at least one of these papers in any resulting publications."), there's no statement requesting astronomers to acknowledge Talk Collections in any paper they publish which makes use of such data.

*there's lots more, in the public domain, on the apparent 'free to raid/use' Talk, by scientists talking with fellow scientists

Posted November 8, 2014 6:35 PM
by vrooje admin, scientist in response to Capella05's comment.

I agree it's always polite to say "I found this in @wtaskew's collection (link to collection)" no matter what. It also makes it easier to trace things later, so that everybody can get credited later should something end up being the focus of a scientific paper or report.

One of the things we've been talking about at ZooHQ recently is the difference between the curated collections and the automatic, hashtag-based collections. There seems to be general agreement that we should try to make the differences clearer in the future, perhaps just by renaming the search-based collections as "Saved Searches" instead of collections.

There are also collections that are in between the two. When you mark a galaxy as a favorite, it gets saved into your Favorites collection, but even though it's a curated collection, favorites aren't necessarily collected according to a theme, and users may not even realize this is happening until after they've been to Talk at least once.

I should also say that I agree that classifying is obviously a really useful and highly valued contribution! 😃

Posted November 8, 2014 6:35 PM
by Capella05 moderator in response to JeanTate's comment.

@JeanTate

How do you know the collection you are raiding is not data for a students term paper? Or a fellow CS long term project (for example the 'Irregulars' project)?

I don't, and can't (unless they say so in their Collection write-up). What's in Talk - Comments, Discussion threads, Collections - is in the public domain. As far as I know, there are no limitations on access to it, and in one sense zooites are more restricted than those who have not registered (non-zooites did/do not have to agree to anything, re use of anything they find here).

That is why I always take the viewpoint that permission has not been granted, and revise my responses accordingly.

Posted November 8, 2014 7:06 PM
by Budgieye moderator

I suppose there are several levels of collection.

The person who classifies, then presses Talk to say "This galaxy is a funny colour"

The person who identifies that the funny colour is Something Interesting, and not just an artifact, and gives the object a tag.

The people who start using the tag to describe similar objects.

The people who create collections of the tagged objects.

The person who skims these collections to make collections of their own.

All these people are necessary because probably no Zooite has classified every galaxy, and at some stage, all should be thanked. I will try to remember the next time I look in collections.

Posted November 9, 2014 6:26 AM
by Budgieye moderator

Should we have a "counter" for each collection. eg. Elisabeth B's lens collection was viewed November 1 by Budgieye.

Posted November 9, 2014 6:28 AM
by Capella05 moderator in response to Budgieye's comment.

That is a nice idea Budgie 😃

Posted November 9, 2014 9:48 AM
by Alpha_Aurigae in response to vrooje's comment.

Hi Brooke,

I am going to try to respond to a number of comments here. Just my opinion, my view and when thrown in with a number of other people's ideas maybe beneficial to Galaxy Zoo as a whole.

I am tagging lots of NGCs, UGCs, MRKs, and other stuff. All those things I find when either classifying galaxies, or checking out volunteers' collections. I don't think there is any harm in that.

I am also on the lookout for interesting galaxies and features, some of them have been identified by other volunteers, some of them not. If they have been mentioned before those volunteers should get the credit, I shouldn't, I just think highlighting an interesting galaxy again to the scientists might draw their attention again to the object (see AGZ0002ja5 for example ).

A couple of days ago I checked out some possible polar rings, and planetaryscience's collection of "Tagged Polar Rings" was very helpful.

I can understand why some people might get uncomfortable when another volunteer checks out their collections. But I want to make one thing perfectly clear: I am not here to steal anyone's find. I love being on Zooniverse and I like nothing more than to contribute to astronomy. When I was on Planet Hunters and started a discussion most of the time I mentioned the volunteer in whose collection I found a certain lightcurve. I wouldn't like someone stealing my find, but I also don't like people stealing other volunteers' finds. I also asked Robert Szabo to include a number of PH volunteers as co-author on the new RR Lyrae paper he is working on. Credit where credit is due.

And yes, I think it is very important volunteers should get recognition or credit for their find (s). I am very adamant about that.

Sometimes it is very clear when someone should get the credit for a discovery. Sometimes not so. I remember zutopian starting a discussion on Planet Hunters some time ago.

Let's take an example. Imagine there is a second Voorwerp on Galaxy Zoo, almost identical to Hanny's Voorwerp.
Who gets the credit?

The volunteer who recognizes it as such but only clicks "other" without leaving any comment?

The volunteer making a comment?

The volunteer who doesn't recognize it as such, but puts it in one of his/her collections anyway?

The volunteer who recognizes it, and makes a comment and either puts it or doesn't put it in a collection?

And then there is the question who is/was first to recognize the second Voorwerp.

Difficult choices and I am glad I don't have to make any decisions in that regard.

I didn't expect this would develop into the discussion it has become. Let's hope we can all benefit from each other's contributions.

And one last time: I am not here to steal any discovery from other volunteers 😃

Posted November 9, 2014 11:36 AM
by Alpha_Aurigae in response to Capella05's comment.

I have to read the other comments elsewhere in this discussion yet (I am reading and commenting on them one at a time), but should there be a Galaxy Zoo version 3 I think it would be useful if the volunteer was given the choice on whether to keep his/her collections public or private.

Everyone can access each other' s collections now, either by clicking on the volunteer' s username, or by clicking the collection next to the galaxy on the star's TALK page. And without making a judgment on whether checking them out is ethical or not, they are public, and they can be inviting, either to check them out for discoveries, or learn from them, or tag galaxies, or use them for your own collections. No one prevents you from checking them out, and also no one doesn't prevent you from checking them out.

I hope this discussion can lead to some useful contributions.

Posted November 9, 2014 11:51 AM
by Alpha_Aurigae in response to JeanTate's comment.

I hope the scientists mining the data are our GZ scientists. Should our data be mined by outsiders they should inform GZ/Zooniverse and give credit where credit is due when something of importance has been found. On Planet Hunters the volunteers do not take kindly to scientists mining the data without ever giving credit, and I am one of them.

But you raise an interesting point. A lot of volunteers put objects in their collections without ever commenting upon them. So you never know WHY they collected the galaxy in the first place. Should there ever be an important discovery, that might become a small problem.

Posted November 9, 2014 12:00 PM
by Alpha_Aurigae in response to Budgieye's comment.

That is a great idea, Budgieye. I favor the counter.

Posted November 9, 2014 12:27 PM
by Capella05 moderator in response to Alpha Aurigae's comment.

Hi Abe,

I am not denying that it is easy to view other volunteer collections, and yes it could be tempting to have a look 😃

I think it is important to point out that the posts and discussions posted in this thread are not specifically aimed towards you, a healthy debate arose from a post you made, and I think I made it clear I was speaking in a personal capacity. I, for one, am happy that this discussion is taking place as it would be interesting to gauge the opinion of other volunteers with regards to the privacy of their collections.

I will be honest I am sometimes stumped as to the purpose of user collections - are we just collecting images that are nice. to look at (much like a stamp or baseball card collection) or do they have a greater purpose? I feel these questions need to be debated before the new version of Talk is developed, and clarification given to all of us.

Without revisiting whether it is ethical or not to view other volunteers personal collections, there is another point that I feel requires discussion. Once again, this is not geared towards you, but rather me, asking a question to the community at large.

If we all spent our time going though each other collections, who is going to do the actual classifying? I know classifying can be tedious at times, and it is tempting to skip that step by just viewing the images on Talk, but all the science that keeps the Zooniverse running is based on our clicks.Will we eventually get to the point that the classifying pool is so small that no new images get posted to Talk? I will admit, a bit over dramatic, but I wanted to stress my point. Classifying is important!

Once again my personal opinion 😃

Posted November 9, 2014 1:06 PM
by Alpha_Aurigae

Hi Julianne,

I am glad we are having this discussion and like I said before I hope something good will come of it (a privacy/public button in the next version of GZ?).

I also hope more volunteers read this discussion, as I value their input as well. Or maybe we should start a new discussion about having private and/or public collections and the ethics of others having a look at them.

I do classify, but the past few months I am having great fun collecting and tagging NGCs, UGCs and MRKs, It also gives me the opportunity to enjoy beautiful images and collect interesting features/pictures and comment on them, e.g. check out the Orion Nebula discussion. And though not a galaxy I think that one is especially interesting and educational for newbies. And maybe experienced volunteers a well. I was quite surprised by how many images there are of the Orion Nebula. And finding the North America Nebula and Pelican Nebula was great as well 😃

I will continue tagging galaxies and studying collections, BUT I will also give credit where credit is due, no one needs to worry I am stealing a find. And if any of the moderators think I am stepping out of line, please say so 😃

This discussion has given me lots to think about and to consider 😃.

Posted November 9, 2014 3:18 PM
by vrooje admin, scientist

I think we'd have to really explore what the implications of a view counter would be for both the experienced and the new volunteers; I'm not convinced everyone would find that a positive experience, but it's worth talking about. It seems like it's definitely in the spirit of giving each volunteer more information and I'm generally really in favor of that.

However, I'm also pretty sure implementing a view counter would be a substantial amount of developer effort, so it's worth submitting that as a suggestion to @DZM -- but realistically I wouldn't expect it to appear in this version of Talk.

I hate it when I find myself being the pessimist. 😦 Apologies.

Posted November 10, 2014 3:32 AM
by JeanTate in response to Alpha Aurigae's comment.

Some really good discussion going here; thanks for kicking this off, AA! 😃

A lot of volunteers put objects in their collections without ever commenting upon them. So you never know WHY they collected the galaxy in the first place.

At some point it may be a good idea to get some data. Based solely on my experience in RGZ, a large majority of objects I found in Collections had comments on them. Sometimes the comments were by the zooite in whose Collection I found the object, sometimes not. As there's no timestamp on when an object is added to a Collection, it is impossible to determine precedence (although with a lot of hard work you could possibly develop a rough timeline).

Also, and this isn't the least bit surprising*, I found that perhaps the majority of objects were in more than one zooite's Collection(s)

So, objects which are in just one Collection and which have no comments^ are rare.

*objects which are collected are frequently unusual; what's unusual to one zooite will often be unusual to many

^or in a Discussion thread; however, there are, from memory, very few of those (though one zooite likes to start Discussion threads rather than write a Comment, I guess for her/him 140 characters is not enough)

Posted November 10, 2014 4:57 AM
by JeanTate in response to Budgieye's comment.

I think ttfnrob (Orbiting Frog) has some data/a paper with numbers. I don't recall whether it includes data on Collections, but I think it mentions that something like ~90% of zooites do not write anything (Comment or post in a Discussion), and of the ~10% who do, ~90% write only Comments (hashtags are Comments). And this was a common pattern across several Zooniverse Talks. (@Capella05: I don't see this behavior changing any time soon! A large majority of zooites will simply classify, as they have always done).

Keep in mind too that because the present Talk's hashtags, Search, and Collections subsystems all have bugs* (they certainly do not work as intended!), there are people (zooites) who try all sorts of things to try to compensate. For example, I began trying to put all the candidate SDRAGNs into a single Collection, but it quickly became very difficult to use, so now I keep track of interesting objects in my own database.

*not all of them known, I'll bet

Posted November 10, 2014 5:11 AM
by JeanTate in response to Alpha Aurigae's comment.

In some ways, waveney's Irregulars Project is like private Collections.

A lot of zooites willingly and gladly participated in that project, and a good deal of very valuable data was collected. However, as far as I know, only waveney has access to it, and he has never published anything on it.

Mentions of ethics reminded me of this: "It is just "I've clicked and been thanked" though." (source). It would seem that whatever you do, or do not, take kindly to doesn't much matter.

Posted November 10, 2014 5:22 AM
by Alpha_Aurigae in response to JeanTate's comment.

Yes, most images I viewed are in more than one collection. Should there ever be a public/private button in a new version of Galaxy Zoo, most information can still be accessed, as not all volunteers will choose to keep their collections private. I am in favor of such a button, it will solve a lot of things. Personally I think I will keep my collections public should I ever been given a choice.

Maybe we should also talk about the search engine on GZ. It is there for a purpose and with it you can easily access comments, collections and discussions. So Galaxy Zoo and maybe Zooniverse makes it very easy and probably inviting again to study each other's collections et al.

A public/private button will solve the matter of having or not having access to other volunteers' collections, but we still see their comments and their participation in discussions, which is also a good thing, because they need to be public.

Posted November 10, 2014 12:41 PM
by DZM admin

Well, a lively discussion broke out over the weekend, didn't it? 😃 Good!

So, I've read through the comments here, and I'm working on trying to understand the issue as best I can... I admit that I'm still trying to grasp the question that was asked earlier in the thread, namely, what are collections really for? It's possible that this has never been completely answered.

With the status quo, where collections are completely public, what is the fear/issue? Is it that someone is going to "steal" someone else's work? It is all about credit for discoveries?

It seems to me that this discussion ties in closely to our very identity as an open-science project, and what that means.

Posted November 10, 2014 3:12 PM
by Capella05 moderator

Sorry I am so late in replying - this is going to be a very hectic week for me, so I will apologise ahead of time for my absence.

Yes, I am not surprised that images are in multiple collections - I would be more worried if they weren't!

I think I am looking at it from a slightly different perspective - perhaps it is not the image itself that is important but rather the combination of images added to a specific collection. Lets look at the following hypothetical situation:
- User A starts to see a unusual pattern in the galaxies s/he has been classifying (not that far fetched, given the Green Peas, PeterD's blue catapillars etc...).
- User A starts a collection to keep track of all of the images that seem to fit the same criteria, but is not ready to publicly discuss it or bring it too the attention of a scientist. Perhaps they wish to collect more data, or research the items themselves before approaching the Science team?
- User B comes along sees that User A is on to something and 'appropriates' the research.
I am not saying that anyone on the Zooniverse would intentionally do that, but the internet is an interesting place 😉

Most scientists will not make the data they have collected open source until such time that they have analysed it and written a paper.

I am not saying that all collections should be private, but surely we should have the option to choose when we make it pubically available?

Posted November 10, 2014 6:35 PM
by Alpha_Aurigae

I am all for that, Julianne! That is why Iike the public/private button 😃.

Maybe the start of a collection could be date-stamped as well?

On Planet Hunters it was no problem checking out volunteers' collections, as far as I know. And when one of us made an interesting find, a number of other volunteers jumped in with their expertise and knowledge to help out (sometimes I have been bit a bit too eager to help out, and I had to be set straight.....).

In case of a very IMPORTANT discovery on Galaxy Zoo, one which should not be made public too soon, the science team and moderators could always choose to hide the discovery from public view and discuss the find with volunteers behind the scenes. IIRC that has been done once on Planet Hunters, but I am not sure about that.

OK, hiding a discovery for a period of time may also raise some questions about ethics, and I haven't given this much thought, but if it would be for the greater good....

Posted November 10, 2014 8:46 PM
by Capella05 moderator in response to Alpha Aurigae's comment.

We are not talking about a discovery per se - but rather giving a volunteer the opportunity to rationally collate and analyse the data they had collected without interference.

If I had been** methodically classifying images and building up a collection with the intention of contacting the science team, it might be somewhat disconcerting to have (without invitation) several volunteers suddenly take it upon themselves to 'investigate' my collection. Those volunteers may feels they are assisting 'User A', but s/he may feel they are trying to take over there work. Remember we are not talking about a user posting a comment - that is open source, but rather creating a collection that they do not necessarily realise is available for everyone to view.

Everyone who makes up our on-line community will have a different idea of 'personal space' and we need to respect that.

Yes, certain volunteers may prefer to go public at the beginning and get the community involved - others may want to do more investigation and approach the Scientists privately. I am all for getting the community involved, but I would first want to take certain investigative steps to make sure what I thought I had found, actually had some scientific merit.

As for "IMPORTANT" discoveries I think GZ has always been very open about them and has certainly never tried to hide anything. Recent case would be 'The Swan'.

Perhaps we are viewing this from two different aspects - I am passionate at giving Citizen Scientist's the opportunity to do more complex tasks (eg. SpaghettiLens) within the mainframe of the Zooniverse, and being able to present there results to a larger community. Recognition is not that important to me.

**I am not, but I am trying to view this debate from all angles.

Posted November 10, 2014 9:57 PM
by Alpha_Aurigae

Again I agree with you Julianne, about giving a volunteer the opportunity to collect galaxies or features in private for study or whatever reason.

And again: a public/private button would come in very handy 😃

Posted November 10, 2014 10:20 PM
by Capella05 moderator

No worries 😃

I hope the debate continues! Now work needs my attention, but I will try to pop in later in the week 😃

Posted November 10, 2014 10:26 PM
by JeanTate in response to Capella05's comment.

Something I've wondering about: isn't this what ZooTools was supposed to be for (at least in part)?

Of course, ZooTools may not have the functionality you need for your particular project, but then it's surely got more and better (bugs etc aside) than Collections does, doesn't it? Or am I misunderstanding something? Perhaps only educators can use ZooTools (but I thought it was available to any zooite)?

More generally, perhaps it might make more sense to tweak ZooTools (and promote it more!) to cater for desires for privacy in Collections, rather than try to shoehorn privacy (etc) into a feature that was never intended to cater for it? I mean, the whole Talk infrastructure is built around being open (as DZM said, the default is open science), while ZooTools is quite the opposite.

Posted November 11, 2014 1:14 PM
by JeanTate in response to DZM's comment.

what are collections really for?

Christian+ (2012) doesn't explicitly address each of Talk's features (so there's nothing there on Collections, specifically), but taken as part of the whole, what Collections are for seems fairly clear (here "tools" means Talk):

The ultimate goal of such tools should be to bring questions and interesting discoveries to the scientists' attention only when expert input is necessary, reducing the time needed for appropriate mentoring while still ensuring nothing gets lost.

One paraphrase might be: Collections are there to increase the efficiency of scientific discovery by allowing professionals to focus on filtered results.

Posted November 11, 2014 1:29 PM
by JeanTate in response to DZM's comment.

[DZM It seems to me that this discussion ties in closely to our very identity as an open-science project, and what that means.

[Capella05] Most scientists will not make the data they have collected open source until such time that they have analysed it and written a paper.

[Alpha Aurigae] In case of a very IMPORTANT discovery on Galaxy Zoo, one which should not be made public too soon, the science team and moderators could always choose to hide the discovery from public view and ...

Three quotes - and there are quite a few others I could have chosen - which point to a larger context; namely Open Science (as DZM noted). It's a huge topic, and we - zooites, Zooniverse projects, etc - are in the middle of it, for better or worse.

Some professionals - e.g. David Hogg - even put draft code up for all to see^ (and download, etc), not to mention all data. Others are more traditional, keeping everything very close to the vest, even after publication. As I understand it, one of the CSA's major backers is strongly in favor of open science (the Adler); perhaps some others are not so enthusiastic?

One aspect of a 'private' button for Collections that might be less than desirable: what happened to the Irregulars Project data*. To avoid data being locked and potentially permanently lost, maybe all Collections marked private will automatically become public after a year? This would be in line with a common practice in astronomy; for example, members of a survey team have access to the data for a year (or some other fixed period) after which it all becomes public.

^shudder; draft code! is there anything more ugly?

*to be clear, this was not a Zooniverse project, and as far as I know the current owner of the data never promised that it would one day be made public

Posted November 11, 2014 1:49 PM
by DZM admin

The way that I see it, Zooniverse is an open-science project. We're trying to move away from that outmoded model where scientists "kept everything close to the vest", because that resulted in a much slower process of discovery. To me, part of the philosophy of Zooniverse is that scientific progress moves faster when the science is done in the open, even if it's done by citizens and volunteers, and not when people are hiding their work in cloistered virtual cubicles.

It's a huge step in the development of a more open scientific community that scientists have started agreeing to release all of their data to the public for projects like those on Zooniverse. I'm very hesitant about anything that would move in the other direction of locking away one's work. It seems to go against the concept of open science.

Collections help the scientists, but they also help the citizen scientists, too! How much easier it would have been to solve the green peas if we'd had collections back then, right?

Posted November 11, 2014 3:01 PM
by Capella05 moderator in response to DZM's comment.

The way that I see it, Zooniverse is an open-science project. We're trying to move away from that outmoded model where scientists "kept everything close to the vest", because that resulted in a much slower process of discovery. To me, part of the philosophy of Zooniverse is that scientific progress moves faster when the science is done in the open, even if it's done by citizens and volunteers, and not when people are hiding their work in cloistered virtual cubicles.

I think things are getting a bit confused here and we are all talking about different things!

I, and I believe the others Zooites that have posted in this thread, are specifically discussing the 'Do-it-ourselves' projects and privacy issues relating to our user collections. We are not referring to how the Zooniverse works 😃

Just as the scientists have the opportunity to analyse the results of our clicks and then present the results, surely we should have the same opportunity when we conduct our own independent research on here?

We are not talking about locking away research, we are talking about giving a volunteer time and perhaps the tools to collect / prepare their findings (in private if they so wish) before releasing it for peer review by the wider Zooniverse community? I am just wary if we did not have the opportunity to conduct this step we may land up doing ourselves more disservice by 'crying wolf' at every opportunity.

Posted November 11, 2014 6:03 PM
by Alpha_Aurigae

Jean Tate: "One paraphrase might be: Collections are there to increase the efficiency of scientific discovery by allowing professionals to focus on filtered results."

That is one of the reasons why I keep collections. I also collect to educate myself, to study objects, and to make an inventory of certain objects like nebula, comets and globular custers. I hve been here now three or four months and after one month I felt the need to collect, and make inventories, just to see what has been happening on GZ before my time. Of course I have been here before, but now that my work on another project is on hold, I intend to make the most of it on Galaxy Zoo, and I am having fun in the process 😃

I am all for openness. But I can also see Julianne's position, and if a private/public button could help solve ethical questions, I say go ahead.

But every volunteer knows these data are on the internet, for every one to see. Now suppose a volunteer found something special, something he doesn't want the others to know about yet, but he is doing more research. In his case I would just copy the url address of the galaxies I was working on and make a list. Click the link and you are back on Galaxy Zoo.

When I was on Planet Hunters and found an interesting lightcurve, sometimes I would hide it in a collection until I had done my research, and could post it on TALK.

There is one other thing that hasn't been mentioned yet.

Say I am collecting with a very special purpose. I found something , and I know astronomers would be interested in that find. Why would I keep that to myself? Isn't our science team here to help us and investigate our find, and make it a genuine Galaxy Zoo find?
But if I were a professional astronomer and use Galaxy Zoo data for myself, and make my collections private, wouldn't that defeat Galaxy Zoo's purpose? We have got very skilled moderators and a very skilled science team. THEY are the first to be addressed, The data may be public but our science team does have a function.
I can tell you over on Planet Hunters they don't take kindly to using the PH data for yourself and not recognizing PH in the process.

While typing this I suddenly think about what it would mean if we really had a public/private button. Wouldn't this invite people to do their own reasearch, and publicize their finds without acknowledging Galaxy Zoo? Sort of defeats Galaxy Zoo's purpose. This is CITIZEN Science after all. And that implies public data.

Posted November 11, 2014 6:27 PM
by DZM admin

I wanted to add: It was mentioned earlier in this thread that our team here, in developing the next Talk, need to be sure that we understand exactly what purpose we want Collections to serve.

That is a conversation that is going on here as we're starting to put together preliminary stuff.

Posted November 11, 2014 8:05 PM
by Alpha_Aurigae in response to DZM's comment.

I think that all depends on the volunteer. I mentioned my reasons in the previous post. But I can image other people collect for different reasons.

And also the purpose of collecting and your collections can change. If for example a volunteer collects images of very beautiful galaxies, he might develop an interest in astronomy and start collecting specific galaxies and objects, and change, delete or expend collections.

I can also image new volunteers or volunteers who visit this site not that often, are collecting just for fun, without any specific and/or scientific purpose. I have studied many collections and a lot, especially the favorites collections, don't contain one or two specific galaxies but a mishmash of galaxies, nebulae and features. And I don't mean that in a deragatory manner.On the contrary. The more things are collected the merrier 😃. Every one is free to collect what he or she wants.

Posted November 11, 2014 9:52 PM
by JeanTate in response to Alpha Aurigae's comment.

(also @Capella05 and DMZ): Interesting points and comments, thank you.

Suppose we're all for open science, and that's our default for everything Zooniverse. Suppose a zooite wishes to do independent research using at least some data which you've obtained from a Zooniverse website (etc) - as you've already done, and as waveney did with the Irregulars Project, etc, etc - how best to foster, support, facilitate, etc that (assuming that's something we'd be keen to do)?

If you had a blank slate, and large (but reasonable) budget/resources/etc, what infrastructure/tools/support/features/etc would you ask for? Something like Collections, yes; what else? ZooTools perhaps? Letters?

From the other side, if various support for ordinary zooites' independent research/study were to be available, what limits should be placed on those, in terms of (eventually) making data (etc) used in such work open (e.g. a limited time period for non-open use)?

Posted November 12, 2014 10:28 AM
by vrooje admin, scientist

I find it hard to separate this discussion from the overall discussion of the Zooniverse philosophy, because the function Talk provides is part of that philosophy. Indeed to me it's a very strong part of it, as the Zooniverse team has mindfully and substantially invested in developing Talk.

As Jean points out, it's possible to keep a private collection right now; you just have to use other software to do it. And to collect things you just need a text file and a folder for saved images. Once you start doing analysis you need more specialized tools and there are strong arguments for having a safe space to play with data; that's part of why the Zooniverse's Tools software already has private workspaces built in. So within Zooniverse software, data analysis can be private, but sample aggregation can't. The way I see it we are talking specifically about whether the Zooniverse should support the private aggregation of data samples on its platform.

My gut reaction to this was "no", in part for reasons Darren mentions above. I can see why the option of a private collection is appealing, but I think the disadvantages of making it a platform standard are pretty significant. So far I haven't changed my mind, though I remain open to my mind being changed.

I also think the discussion of whether and how the Zooniverse should facilitate the whole scientific process is important, but that is a very big generalization from where this started -- at least to me.

Cheers,

-Brooke

Posted November 12, 2014 5:32 PM
by JeanTate in response to DZM's comment.

@vrooje: we have, indeed, begun to move a long way from where this thread began.

Perhaps DZM put his finger on it:

our team here, in developing the next Talk, need to be sure that we understand exactly what purpose we want Collections to serve

So far, the sole intended purpose for a private/public button on Collections that has been proposed has been an individual zooite's own (private) study*. And in discussing this, I can't see how we could avoid considering the broader context.

But perhaps there are other reasons why having a private/public button on Collections might be a good idea? Perhaps a zooite might like to have that option, for purposes other than their own individual research?

More generally, what do zooites use Collections for^? Has there been any research into that question?

*it would have to be just for the individual; once private, only you could view a Collection which you (the creator) had designated "private"

^we have AA's input, upthread: "to educate myself, to study objects, and to make an inventory of certain objects like nebula, comets and globular custers"

Posted November 14, 2014 4:05 PM
by Capella05 moderator

From the T&C's of the Zooniverse:

What you may do with Zooniverse data

You retain ownership of any contribution you make to the Zooniverse,
and any recorded interaction with the dataset associated with the
Zooniverse. You may use, distribute or modify your individual
contribution in any way you like. However, you do not possess
ownership of the dataset itself. This license does not apply to data
about you, covered in the Privacy Policy.

I guess that means that any collection you personally create is under your ownership.

Posted November 21, 2014 11:28 PM
by Alpha_Aurigae

Of course, our collections are our own, we collected all those images, but never does it say anywhere hands off 😃

But you are opening another can of worms, and I think that should be addressed as it hasn't been discussed on Zooniverse before, not at least as far as I know.

I know about this statement you copied. You may freely distribute your individual contribution in any way you like. I know for a fact PH and GZ don't appreciate a volunteer contacting ouside astronomers if the volunteer has made a (significant) discovery. That is why we have science teams. They are the first people to go to. And that is how it should be.

Don't worry, I have no intention to contact any outside astronomers 😃. Should I ever want to I'll contact the moderators and the science team first.

Posted November 22, 2014 9:14 AM
by Capella05 moderator

You missed the point 😃

Ownership = The act, state, or right of possessing something - Oxford Dictonary

By that definition alone, permission (from the owner) would be needed to access / use something that belongs to another.

Once again this is just in reference to the collections debate, nothing else, and it my own opinion.

Posted November 22, 2014 2:07 PM
by Alpha_Aurigae

I think this would a as good a time as any to ask the Zookeeper himself what he thinks.

Nowhere does it explicity say you OWN your collections , but it neither explicitly says you DON"T OWN the collections.There is also no mention of people allowing to use the data in your collections, or not allowing to use those data.

This is going to be a long discussion of yes and no, and everyone is right to a certain point.

I wouldn't mind if Chris Lintott addresses this issue here 😃

Posted November 22, 2014 3:16 PM
by williamaskew

i have been collecting for years and of course I dont own them, I just put them in order so i can access them with ease when I want to view them, or if others feel inclined.I have dozens of collections and thousands of images. Help your self.

Posted November 23, 2014 6:12 PM
by vrooje admin, scientist

I've said before that I doubt any changes discussed here will be implemented the current version of Talk. Also, there is no promise that any features will be implemented in a new version of Talk, even if we did all agree.

As Capella05 said, collections are made up largely of "the dataset itself". The extra information provided in the collection name, description and comments is value added by each volunteer, but no volunteer owns the images themselves.

Talk is also specifically excluded from the confidentiality part of the T&Cs, which makes it pretty clear that Talk is intended to be public: "Contributions you make to the Talk pages are widely available to others."

It is appropriate to give credit to someone's collection when you've found an image in that collection, especially if the collection's stated purpose is in any way related to the reason you're re-posting the image. If a major discovery is made this way, the credit should go at least in part to the person who classified the image and also flagged it as interesting (by collecting it). If someone tells you they would prefer you not search through their collections, it would save us all a lot of trouble if you respected that even though you are not obligated to do so.

One last note: I have heard that if you say zookeeper's name three times when you look in the mirror, he will appear and tell you how many unread messages there are in his inbox.

Posted November 24, 2014 3:15 PM
by Alpha_Aurigae

I tried using the ruby slippers three times and hoped to teleport inside Chris' offices, but alas, I ended up elsewhere 😉

" It is appropriate to give credit to someone's collection when you've found an image in that collection, especially if the collection's stated purpose is in any way related to the reason you're re-posting the image. If a major discovery is made this way, the credit should go at least in part to the person who classified the image and also flagged it as interesting (by collecting it). If someone tells you they would prefer you not search through their collections, it would save us all a lot of trouble if you respected that even though you are not obligated to do so." Now this is exactly how I feel, and it is the best way to handle this. We also did this on Planet Hunters, and people who found a RR Lyrae (sometimes even without knowing it - Even Hanny didn't know she had discovered Hanny's Voorwerp), will be mentioned in the new RR Lyrae paper.

I have always been a strong supporter of giving credit where credit is due, and I am glad you agree 😃 😃 😃

Posted November 24, 2014 3:48 PM
by zutopian

There hasn't been any new post for one year in this discussion. I do a post, because I want to inform, that there is following new discussion, which is related.:

'Favourites' dataset?
http://talk.galaxyzoo.org/#/boards/BGZ0000007/discussions/DGZ0001pkw

Posted November 19, 2015 6:29 PM