[Tf-curation] FW: CLARIN2017 notification for paper 9

[Menzo Windhouwer2]

Hi, All,

Good news! Our abstract on the CMDI Best Practices guide has been accepted for the upcoming CAC :-)

In August we can work, if you’re not on holiday ;-), on the final version. This nicely lines up with the plan to have a first (draft) version of the guide as a CE document.

See you soon at the CMDI BP vidconf (21 July 10:00) or have nice summer break!

Menzo
--
www.meertens.knaw.nl/cms/medewerkers/144709-menzowi

On 02/07/2017, 20:40, "CLARIN2017" <cac2017-0 at easychair.org> wrote:

    Dear Menzo,
    
    On behalf of the CLARIN Annual Conference 2017 Program Committee, it is great pleasure to inform you that your submission:
    
    "Component Metadata Infrastructure Best Practices for CLARIN"
    
    has been accepted to be orally presented at the conference.
    
    
    The Program Committee worked very hard to thoroughly review
    all the submitted papers.  Please repay their efforts, by 
    following their suggestions when you revise your extended abstract.
    
    The deadline for uploading the camera-ready version of the extended abstract is September 1st, 2017. Please, use the CLARIN 2017 Easy Chair account for this purpose:
    https://easychair.org/conferences/?conf=cac20170
         
    All authors of the accepted abstracts will be invited to submit full versions of the papers to the post-conference proceedings after the conference. A separate call will be distributed after the conference.
    
    The reviews and comments are attached below. Please, consider them carefully and adapt the final form of the extended abstract, as well as the presentation according to them.
    
    We would like to remind you that there is no quality difference between
    the oral and the poster presentations. The choice of style of presentation 
    of your paper was made based on what would be the most informative and 
    effective way to communicate the scientific contents of your proposal.
    
    For each accepted paper exactly one author will be granted reimbursement of travel costs (up to 220 Euros), and free accommodation and meals. Please, contact as quickly as possible the National Coordinator of the CLARIN member (country or organisation), which is appropriate for the author who will be presenting the work and attending the conference.
    
    
    Congratulations on your fine work!
    
    
    Best regards,
        Maciej, on behalf of the CLARIN 2017 PC
    
    
    ----------------------- REVIEW 1 ---------------------
    PAPER: 9
    TITLE: Component Metadata Infrastructure Best Practices for CLARIN
    AUTHORS: Thomas Eckart, Twan Goosen, Susanne Haaf, Hanna Hedeland, Oddrun Ohren, Dieter Van Uytvanck and Menzo Windhouwer
    
    Relevance: 5 (excellent)
    Content: 4 (good)
    State of the art: 3 (fair)
    Implementation: 3 (fair)
    CLARIN impact: 4 (good)
    Language and structure: 4 (good)
    Overall evaluation: 2 (accept)
    Reviewer's confidence: 5 (expert)
    
    ----------- Overall evaluation -----------
    This paper, anticipating the release of the CMD Best Practices document for CLARIN, is clearly very much in line with the topics of the conference. CMDI is one of the pillars of CLARIN, allowing for metadata exchange between centers and the infrastructure. 
     
    The paper opens up with a sentence whose importance that cannot be stressed enough:
    “...design and implementation choices made at various levels in the CMD lifecycle might influence how well or easily a CMD record is processed and its associated resources made available in the CLARIN infrastructure.”
     
    Clearly we cannot appropriately judge the work carried out by the authors based only on the 4 pages of the abstract, and we shall need to wait for the full document. However, I am deeply convinced that the discussion on these issues is very important, and this partly in disagreement with what the authors say in the paper: in fact while the final document will mostly address the categories they mention, I believe that a larger discussion on topics such as searchability and facet mapping, granularity and representation of hierarchies have a scope that goes beyond the close circle of curation experts and strategically involves a broader audience within the national consortia. For instance, in my personal experience, one of the main inducements for LR owners to deposit their data in a CLARIN center is to enhance the visibility of their work. It is important for them to know that the implementation of profiles in national repositories may have an impact on this.
    More broadly, these choices may impact on how well the resources interact with tools such as Weblicht or the Switchboard.
    Finally, and crucially we cannot be blind to the fact that many CLARIN repositories adopt solutions that may limit the possibility of implementing some of the best practices mentioned here. 
     
    For these reasons I believe we need a general discussion on this, and the CLARIN conference is a good place to start.
     
    Some remarks:
    - It would be great if at least a pre-final draft of the guide were available at the time of the conference
    - For uniformity to other papers, a short abstract may be added
    - The paper and the presentation could be more explicit as to which is the current state of affairs concerning such best practices in various centers, and as to which are the foreseeable obstacles to their widespread application
    
    
    ----------------------- REVIEW 2 ---------------------
    PAPER: 9
    TITLE: Component Metadata Infrastructure Best Practices for CLARIN
    AUTHORS: Thomas Eckart, Twan Goosen, Susanne Haaf, Hanna Hedeland, Oddrun Ohren, Dieter Van Uytvanck and Menzo Windhouwer
    
    Relevance: 5 (excellent)
    Content: 4 (good)
    State of the art: 5 (excellent)
    Implementation: 3 (fair)
    CLARIN impact: 5 (excellent)
    Language and structure: 4 (good)
    Overall evaluation: 3 (strong accept)
    Reviewer's confidence: 4 (high)
    
    ----------- Overall evaluation -----------
    This abstract draws our attention to the Best Practice Guide that is to appear in the course of 2017 and illustrated some of the recommendations and guidelines it will contain.
    
    The content is highly relevant to CLARIN and very much demanded. The paper shows excellent knowledge of and expertise in the matters by the authors. The work has not been finished yet but will probably be finished by the time of the CLARIN Conference.
    
    The abstract is quite general and in various respects very ‘abstract’. Concrete examples of existing metadata records that violate some of the recommendations and suggestions on how the metadata can be improved by sticking to the recommendations would be very helpful, and will also justify formulation of the guidelines.
    
    Among the guidelines I miss a list of minimally required or strongly recommended metadata elements (defined semantically), although the VLO facets are explicitly mentioned. It is my experience that researchers who make metadata for the resources they work with leave out information about their resources that are obvious to them and not distinctive for the set of data they work with. Thus records have appeared without an indication of resource type (because the researchers involved only work with text corpora) or language (because the researchers involved always work with one and the same language), etc. Very often, researchers do not assign a name to a resource, let alone a version number. CLARIN should guide these researchers, so that such omissions are explicitly pointed out to them (preferably by automatic checks) so that if they deviate from it they do so consciously and with good justification rather than out of lack of knowledge or accidental omission. It is also impor!
     tant to point out that the profile name itself will not be part of the metadata: I have seen quite some metadata profiles for a specific resource type X that never explicitly state that they describe a resource of resource type X because the profile name contains a label for resource type X.
    
    Minor remarks:
    Sec 4.2.2 In addition, ….should be attended to. :  this sentence is incomprehensible
    Sec 5 : joined -> joint
    
    
    ----------------------- REVIEW 3 ---------------------
    PAPER: 9
    TITLE: Component Metadata Infrastructure Best Practices for CLARIN
    AUTHORS: Thomas Eckart, Twan Goosen, Susanne Haaf, Hanna Hedeland, Oddrun Ohren, Dieter Van Uytvanck and Menzo Windhouwer
    
    Relevance: 5 (excellent)
    Content: 4 (good)
    State of the art: 3 (fair)
    Implementation: 4 (good)
    CLARIN impact: 4 (good)
    Language and structure: 5 (excellent)
    Overall evaluation: 2 (accept)
    Reviewer's confidence: 5 (expert)
    
    ----------- Overall evaluation -----------
    The CMDI Best Practice Guide will prove valuable for CMD modellers and authors. It seems to be relying on an evaluation of the current situation and tries to adddress some of the known issues. It is not clear from the paper whether at least some aspects/pieces of the best practices will also be technically supported/enforced.