Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[bioontology-support] Old NCBO Annotators parameters ?

manuelso manuelso at stanford.edu
Tue Nov 26 09:07:03 PST 2013


Hi Clement,

Missing documentation will be out in the next release.

See other comments below …

Manuel

On Nov 26, 2013, at 4:54 AM, Clement Jonquet <jonquet at lirmm.fr> wrote:

> Hi Manuel,
> Thanks for your responses.
> See some remarks in green.
>  
> Clement
>  
> De : manuelso [mailto:manuelso at stanford.edu] 
> Envoyé : jeudi 21 novembre 2013 18:22
> À : Clement Jonquet
> Cc : support at bioontology.org Support
> Objet : Re: [bioontology-support] Old NCBO Annotators parameters ?
>  
> Hi Clement,
>  
> I think some of your points here come from incomplete documentation that we need to fix. Thanks for pointing these out.
>  
> There are a couple of questions that need investigation, the ones with tickets created. I will respond to them once we learn what the solution is.
>  
> Find answers inlined below.
>  
> Manuel
>  
>  
> On Nov 18, 2013, at 12:18 AM, Clement Jonquet <jonquet at lirmm.fr> wrote:
> 
> 
> Hi all,
>  
> Will the old NCBO Annotator parameters :
> longestOnly
>  
> This parameter was removed due to inability to perform longestOnly on a set of ontologies. mgrep does longest work across the entire dictionary.
> 
> è Can you take the occasion of the new release to have “one mgrep per ontology” as it was discussed a few years ago ? 

I will mention this to the team and push it as a feature request. I have created a ticket for it.

> Mgrep is intentionally designed to be run with several daemons in parallel (see doc).
> This feature is quite missing for anyone who want to reduce the number of annotation returned by a picece of text.
> (We have e.g., someone at LIRMM (Sandra) who is interested in matching the longest term only).
> wholeWordOnly
>  
> We do not have this flag at the moment and I am not sure of what the expected behavior is here. I have created aticket to investigate this.
>  
> è Not documented at : http://data.bioontology.org/documentation#nav_annotator
>  
> filterNumber
>  
> became exclude_numbers
>  
> è Not documented at : http://data.bioontology.org/documentation#nav_annotator
>  
> minTermSize
>  
> became minimum_match_length
>  
> è Yep, sorry I forgot to remove this one when wrote the email ;)
>  
> withSynonyms
>  
> By default include synonyms but I believe we do not have a flag to disable this at the moment. I have created a ticket to address this issue.
>  
> è Ok, thanks Could be useful maybe.
>  
> come back sometime in the new implementation of the Annotator ?
>  
> Also, what happen to the  Scored parameter, and more important, is there any more scoring ?
>  
> The scoring was removed. I all ask Ray to comment on this.
> 
> è Ok. If you can provide more information about this then I am interested. By the way, I a am working with Juan a student of mine using an NLP measure that I am thinking about using to enhance scoring (based on the term embedded in longer terms and frequency). I can discuss this off list you you like.

Sure. Contact me any time.

>  
> This was quite important to distinguish annotations one another and to aggregate them all.
>  
> Also, is the meaning of the new ‘ontologies’ parameter
> ontologiesToExpand or  ontologiesToKeepInResult ?
>  
> The ontologies param aligns with the old ontologiesToKeepInResult. To expand one needs to use the class endpoints.
>  
> I don’t understand the second part of the response, could you elaborate a bit more ?

Sorry my answer was cryptic indeed. I meant that to expand a term with mappings one needs to use the mapping REST services.

> 
> 
> i.e., are you using all the ontologies to expand with mappings and then keep only the results in the ‘ontologies’ parameter ? (remember the “treatment” & “therapeutic procedure” example.
>  
> Bye
> Clement
>  
>  
> ----------------------------------------------------------------------------------------------------------------------
> Dr. Clement JONQUET  -  PhD in Informatics  -  Assistant Professor
> PI of the SIFR project: Semantic Indexing of French Biomedical Data Resources (ANR, UM2, CNRS)
>  
> jonquet at lirmm.fr
> http://www.lirmm.fr/~jonquet
>  
>                 University of Montpellier                          Tel:                      +33/4 67 14 97 43
>                 LIRMM                                                            Fax:                     +33/4 67  41 85 00          
>                 161 rue Ada                                                   Skype:                 clementpro
>                 34095 Montpellier Cdx 5                            Twitter:               @jonquet_lirmm
>                 France                                                            Google:               jonquet.lirmm at gmail.com
> ----------------------------------------------------------------------------------------------------------------------
> <image001.png>
>  
>  
>  
>  
> _______________________________________________
> bioontology-support mailing list
> bioontology-support at lists.stanford.edu
> https://mailman.stanford.edu/mailman/listinfo/bioontology-support
>  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.stanford.edu/pipermail/bioontology-support/attachments/20131126/706a3ea5/attachment-0001.html>


More information about the bioontology-support mailing list