Re: morph module development issues

Dear Bettina,

I think the poll mechanism is a good idea, if used complementary to  
discussions. My personal feeling is that people are waiting to get  
everything a little bit more precise and explicit to really feel able to  
state their opinion on certain modeling choices, and such a poll mechanism  
could be a good way to speed the decision process up -- even if certain  
decisions may be revoked during later discussions, when other parts of the  
model are becoming clearer. (I have to admit I haven't been the most  
active telco participant either, but our group is always present with one  
person at least, and I tried to comment extensively in the wiki and the  
minutes.)

What I am less sure about is when to use the poll. I think that this would  
only be advisable if telco participants don't agree or really don't know,  
but not for second-guessing possible motivations beyond their personal  
expertise. If a use case or a modelling choice cannot be justified by the  
telco participants, from the shared data samples or by other documentation  
produced beforehand, it should not be considered during the  
discussion/initial design at the telco. Whenever a concrete proposal is on  
the table, it can still be extended and adjusted as needed, and I think,  
more controversy (and better attendance) will come automatically, then,  
but the polls should not replace the discussion.

Best,
Christian

Am .03.2019, 21:39 Uhr, schrieb Bettina Klimek  
<klimek@informatik.uni-leipzig.de>:

> Dear OntoLex community,
>
> during last Tuesday's telco, Max and I were the only participants. As  
> you can conclude, not much progress could be achieved and this  
> circumstance leads me to openly sharing my concerns regarding the  
> development of this module with you. After three months I have to  
> summarize that, unfortunately, not much progress could be achieved. Only  
> very few people are able to participate regularly in the telcos and  
> those are often not lexicographers. Arising issues that need to be  
> discussed by more members of the community group and for which I asked  
> for feedback via e-mail, receive very little to zero feedback.
>
> As a result, we are stuck with many open questions about whether  
> lexicographers have certain kind of morphological data at all or if they  
> recommend it to be representable with the new module. Especially those  
> who participate but are not coming from the lexicography domain are  
> (understandably) reserved to make decisions for the target user group.  
> Within the telcos we are thrown back on either re-explaining  issues  
> that have been discussed before to people who were unable to attend  
> earlier calls or on collecting arguments for or against modelling  
> proposals without reaching any consent.
>
> I do understand that not everybody can attend each and every telco or  
> has the time to follow up on what has been worked out so far. However,  
> without almost no significant input via e-mail, adding comments to the  
> wiki or the minutes documents, I do not see how this module can be  
> developed as a community effort. I am missing clear module development  
> guidelines and procedures for arriving at decisions.
>
> I am also putting my time and effort into leading the development of  
> this module and would like to see reasonable progress accordingly. The  
> only way I see to achieve this is by finding a way to arrive at (at  
> least preliminary) decisions in order to move on to the next issue.  
> Therefore, I propose to work with Google forms that I would prepare and  
> send via the mailing list in order to reach consensus on certain topics  
> that are necessary prerequisites for modelling questions. Such a poll  
> will consist of the outline of the issue under question and a list of  
> all proposed solutions with their pros and cons which can be chosen from.
>
> In order to find out what your opinions are regarding this proposal of  
> using polls I created a Google Form which I kindly ask you to fill in :)
>
> https://goo.gl/forms/cAYwlYCqMwARDhK32
>
> I am looking forward to your responses.
>
> All the best,
>
> Bettina
>
>


-- 
Prof. Dr. Christian Chiarcos
Applied Computational Linguistics
Johann Wolfgang Goethe Universität Frankfurt a. M.
60054 Frankfurt am Main, Germany

office: Robert-Mayer-Str. 11-15, #107
mail: chiarcos@informatik.uni-frankfurt.de
web: http://acoli.cs.uni-frankfurt.de
tel: +49-(0)69-798-22463
fax: +49-(0)69-798-28334

Received on Monday, 18 March 2019 06:35:42 UTC