- From: Eric Stephan <ericphb@gmail.com>
- Date: Wed, 22 Apr 2015 19:41:36 -0700
- To: Laufer <laufer@globo.com>
- Cc: Bernadette Farias Lóscio <bfl@cin.ufpe.br>, "public-dwbp-wg@w3.org" <public-dwbp-wg@w3.org>
- Message-Id: <420C8257-4684-4510-BE5B-060FB177C73F@gmail.com>
Laufer , Take a look at my comments below and let me know what you think. Thanks so much! Cheers Eric Sent from my iPhone > On Apr 22, 2015, at 6:56 PM, Laufer <laufer@globo.com> wrote: > > Eric, > > Your definition of Annotation seems as a type of metadata provided by the producer. For the Web Annotation WG a big set of Annotations are provided by the Readers (Consumers). And these Annotations could also be annotated, creating threads (conversations), that seems with your definition of Feedback. > Interesting. I get the concept of annotation, this is commonly used in scientific collaborations. E.g scientist A adds metadata to scientist's B dataset. Again we need to bring the formal definitions of Annotation into what we are doing. Thank you for sharing. Feedback could be considered another type of annotation, however I like the idea of keeping documentation and feedback separate. Documentation (annotation) can be overwritten but feedback is only appended in a directed graph. Does this sound reasonable? > Maybe our Annotations are really different from the Web Annotation WG. > > Feedback seems to me (just feelings) to be more than the (human) conversations around the Dataset. > > I thought that Citation was the inverse: the reference that other published material makes to the Dataset. Sorry. > Why couldn't citation represent citation e.g DOI for dataset, and what I was proposing. I think we should accommodate both if possible. > Your definition of Citation is to reference who talks about the Dataset. In my definition, Citation is how one references the Dataset when talking about the Dataset. > > Cheers, > Laufer > > > Em quarta-feira, 22 de abril de 2015, Eric Stephan <ericphb@gmail.com> escreveu: >> Laufer, >> >> Many thanks for your "feedback" :-) For reference here are some definitions to show you point of view and comments to your statements are below with the --> >> >> >> Cheers, >> >> Eric >> >> Definitions from my Point of View >> To me the 3 distinct elements of the Dataset Usage Vocabulary are: usage, feedback, and citation. >> >> Usage >> UsageAnnotation is a form of Dataset documentation most likely provided by the publisher or producer. >> DatasetUsage->Application is a form of Dataset documentation that describes an application that can use the data. >> >> Feedback >> Feedback will mostly be a dialog that a data publisher sets up for data consumers to provide comments about the Dataset or to each other. >> >> Citation >> Citation from a Dataset Usage Vocabulary perspective is a reference to other published material about the Dataset. >> >> >> Comments to your questions and concerns below.... >> >>> On Wed, Apr 22, 2015 at 9:39 AM, Laufer <laufer@globo.com> wrote: >>> Hi, Bernadette, Eric, >>> >>> First of all, thank you. >>> >>> I would like to know what are the differences between what you are calling Feedback and Annotation. To me, Annotation is a way to provide Feedback. >> >> --> Based on the way I'm defining it at this point I disagree. >> >>> Second, Feedback has specializations that are of different natures: Opinion and Rating are related to the content type of the Feedback (Annotation?), while Blog is related to the type of the place of the content (that could be an Opinion with a Citation, etc.). >> >> --> I think if we combine Feedback and Annotation we lose the distinction of who said what. >> >>> In the diagram, Citation has no relation to an Agent. An Agent cites a Citation. >> --> I see your point, I agree. >> >>> I don't understand exactly the relation of Citation with DatasetUsage. As I said in a previous e-mail, I see Citation as a relation between the Dataset and an Annotation that provides Feedback. The number of Citations give some kind of feedback (as h-index) but the content where the the Citation is used is very important (the reverse link). >> --> Based on the way I described things about I'm tending to agree with you about the relationship between Dataset Usage and Citation. >> >>> We have to take care with the term Annotation because the Web Annotation WG has a well-defined meaning to that. Are we using the same meaning? >> --> Yes I believe that is what we mean, although we need to look at the way the Annotation working group is defining it to make sure that is the correct way to go. >> >>> Cheers, >>> Laufer >>> >>> 2015-04-22 12:12 GMT-03:00 Eric Stephan <ericphb@gmail.com>: >>> >>>> >>>> Thanks so much Bernadette! My comments below. >>>> >>>> Eric >>>> Sent from my iPhone >>>> >>>>> On Apr 22, 2015, at 7:18 AM, Bernadette Farias Lóscio <bfl@cin.ufpe.br> wrote: >>>>> >>>>> Hi Eric, >>>>> >>>>> Thank you very much for the examples! They are really useful! My comments are inline. >>>>>> >>>>>> Example 1: Jian wants to provide "readme" information that includes recommended uses and some background information with a previously published climate model diagnostics dataset http://example.com/atmos/sgp-uncertainty/ . To do this Jian publishes "duv:UsageAnnotation". >>>>> >>>>> Yes. In this case, Jian creates an annotation that describes a specific type of DatasetUsage (ex: duv:GeneralInformation). However, the current version of the model doesn't have this type of DatasetUsage.Should we include a new subclass? >>>> I was thinking this was UsageAnnotation, no? >>>>> >>>>>> >>>>>> Example 2: Laura has a perl script 2d-plotter.pl that creates quarterly (3 month) jpeg plots from netcdf files contained in the model diagnostics dataset http://example.com/atmos/sgp-uncertainty/ . Laura makes the perl script available in github and publishes "duv:Application" that associates 2d-plotter.pl with the dataset. >>>>> >>>>> Yes. From the point of view of Laura, she publishes a duv:Application. However, from the point of view of the Publisher, it is also possible to associate a duv:UsageAnnotation to the dataset in order to provide information about who is using the dataset. Does it make sense for you >>>> Yes >>>>>> >>>>>> Example 3: Matt publishes a peer reviewed technical report on errors and imprecision found in the model diagnostics dataset over limited time periods when the observational data was collected. Matt publishes "duv:Citation" that associates a citation reference to the technical report. >>>>> >>>>> Yes! This is similar to Example 2. >>>>> >>>>>> >>>>>> Is this how you are thinking it would work? I am beginning to wonder if we really need the DatasetUsage class, I don't know what this adds. It is helpful in the conceptual model, but in practice it seems like it unnecessary overhead. >>>>>> >>>>>> What do you think? >>>>> >>>>> I agree with you that in practice maybe it won't be necessary to have the DatasetUsage class. However, let's wait a little more before to remove it :) Let's see if it is really unnecessary. >>>> >>>> I am fine with this, thanks! >>>> >>>>> >>>>> Cheers, >>>>> Bernadette >>>>> >>>>>> >>>>>> Eric S >>>>>> >>>>>> >>>>>>> On Wed, Apr 22, 2015 at 5:37 AM, Bernadette Farias Lóscio <bfl@cin.ufpe.br> wrote: >>>>>>> Hi Eric, >>>>>>> >>>>>>> The idea of having the class duv:UsageAnnotation is to have a way to describe how datasets can be annotated with information about their usage. In this case, I don't see it as a subclass of duv:DatasetUsage. >>>>>>> >>>>>>> We can also have a similar way to annotate a dataset with information about feedback gathered from consumers. >>>>>>> >>>>>>> I think that in the dataset usage vocab, we should consider two scenarios. The first one is how data consumers will provide information about the dataset usage and feedback. The second one is how a dataset can be annotated with such information. Does it make sense for you? >>>>>>> >>>>>>> Cheers, >>>>>>> Bernadette >>>>>>> >>>>>>> 2015-04-22 9:14 GMT-03:00 Eric Stephan <ericphb@gmail.com>: >>>>>>> >>>>>>>> Bernadette, >>>>>>>> >>>>>>>> I really like where you put data citation as a subclass to duv:DatasetUsage (I changed the class name, do you agree?), it seems like UsageAnnotation could be a subclass to DatasetUsage. What are your thoughts? >>>>>>>> >>>>>>>> Eric S >>>>>>>> >>>>>>>>> On Wed, Apr 22, 2015 at 5:01 AM, Bernadette Farias Lóscio <bfl@cin.ufpe.br> wrote: >>>>>>>>> ok Eric! Thank you! >>>>>>>>> >>>>>>>>> cheers, >>>>>>>>> Bernadette >>>>>>>>> >>>>>>>>> 2015-04-22 8:55 GMT-03:00 Eric Stephan <ericphb@gmail.com>: >>>>>>>>> >>>>>>>>>> Thank you Bernadette, it looks great. I'm just responding with the action number to record your email in the action tracker. >>>>>>>>>> >>>>>>>>>> Cheers, >>>>>>>>>> >>>>>>>>>> Eric S >>>>>>>>>> >>>>>>>>>>> On Tue, Apr 21, 2015 at 6:17 PM, Bernadette Farias Lóscio <bfl@cin.ufpe.br> wrote: >>>>>>>>>>> Hi all, >>>>>>>>>>> >>>>>>>>>>> Now we have an editable version of the DUV model available on google drive [1]. >>>>>>>>>>> >>>>>>>>>>> Please feel free to make comments and updates! >>>>>>>>>>> >>>>>>>>>>> Cheers! >>>>>>>>>>> Bernadette >>>>>>>>>>> >>>>>>>>>>> [1] https://docs.google.com/drawings/d/1aq3vPcoj0SPs5BispD6umQNejrBTwkhsSYu6Y1adUjw/edit?usp=sharing >>>>>>>>>>> -- >>>>>>>>>>> Bernadette Farias Lóscio >>>>>>>>>>> Centro de Informática >>>>>>>>>>> Universidade Federal de Pernambuco - UFPE, Brazil >>>>>>>>>>> ---------------------------------------------------------------------------- >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> -- >>>>>>>>> Bernadette Farias Lóscio >>>>>>>>> Centro de Informática >>>>>>>>> Universidade Federal de Pernambuco - UFPE, Brazil >>>>>>>>> ---------------------------------------------------------------------------- >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Bernadette Farias Lóscio >>>>>>> Centro de Informática >>>>>>> Universidade Federal de Pernambuco - UFPE, Brazil >>>>>>> ---------------------------------------------------------------------------- >>>>> >>>>> >>>>> >>>>> -- >>>>> Bernadette Farias Lóscio >>>>> Centro de Informática >>>>> Universidade Federal de Pernambuco - UFPE, Brazil >>>>> ---------------------------------------------------------------------------- >>> >>> >>> >>> -- >>> . . . .. . . >>> . . . .. >>> . .. . > > > -- > . . . .. . . > . . . .. > . .. .
Received on Thursday, 23 April 2015 02:42:08 UTC