[RNLD] importing/exporting between ELAN and FLEX

[Kate Horrack]

Hi John,

In ELAN, I have my tiers set up as follows:

*Transcription Tier*:

   - Tier Name: *SpeakerName*-TR
   - Parent Tier: None
   - Linguistic Type: TR
   - Default Language: None


*Blank Tier*:

   - Tier Name: *SpeakerName-*BLANK
   - Parent Tier: *SpeakerName*-TR
   - Linguistic Type: BLANK
   - Default Language: None

With regard to types, these are set up as follows:

*TR:*

   - Type Name: TR
   - Stereotype: None
   - Time-alignable: Yes

*BLANK;*

   - Type Name: BLANK
   - Stereotype: Symbolic Association
   - Time-alignable: No


With regard to the export process, I deselect the blank tier in Step 2 so
that it's not exported (underneath the 'phrase_item' column in the first
table). When you deselect this, it just says '<select a type>' instead. In
Step 3, you need to select a language in the third column. To do this, go
down to the box called 'Type-Lang value configuration', select 'language',
type in a value (I use NUY, which is the language code of the language I'm
working on), and click 'Add'. Then, go up to the table, and select the
value that you just added as your language in the third column. I tried to
attach a word file with screen shots showing how I have things
selected/entered at each step, but the email didn't go through (I didn't
realise you can't send attachments with this mailing list). Hopefully you
received it though, as the email was also sent to your own email address.
If not, let me know and I'll try again.

Most of this I worked out through trial and error, so there may be better
ways of going about it. But it works! (That's the main thing, right?)

Cheers,
Kate.


On Fri, Sep 13, 2013 at 8:51 AM, Kate Horrack <
khorrack at student.unimelb.edu.au> wrote:

> Hi John,
>
> In ELAN, I have my tiers set up as follows:
>
> *Transcription Tier*:
>
>    - Tier Name: *SpeakerName*-TR
>    - Parent Tier: None
>    - Linguistic Type: TR
>    - Default Language: None
>
>
> *Blank Tier*:
>
>    - Tier Name: *SpeakerName-*BLANK
>    - Parent Tier: *SpeakerName*-TR
>    - Linguistic Type: BLANK
>    - Default Language: None
>
> With regard to types, these are set up as follows:
>
> *TR:*
>
>    - Type Name: TR
>    - Stereotype: None
>    - Time-alignable: Yes
>
> *BLANK;*
>
>    - Type Name: BLANK
>    - Stereotype: Symbolic Association
>    - Time-alignable: No
>
>
> With regard to the export process, I've attached a word file with screen
> shots showing how I have things selected/entered at each step. I deselect
> the blank tier in Step 2 so that it's not exported (underneath the
> 'phrase_item' column in the first table). When you deselect this, it just
> says '<select a type>' instead. In Step 3, you need to select a language in
> the third column. To do this, go down to the box called 'Type-Lang value
> configuration', select 'language', type in a value (I use NUY, which is the
> language code of the language I'm working on), and click 'Add'. Then, go up
> to the table, and select the value that you just added as your language in
> the third column.
>
> Most of this I worked out through trial and error, so there may be better
> ways of going about it. But it works! (That's the main thing, right?)
>
> Cheers,
> Kate.
>
>
>
>
>
> On Thu, Sep 12, 2013 at 11:52 PM, Jonathon Lum <lum.jonathon at gmail.com>wrote:
>
>> Hi Kate,
>>
>> Thanks for your reply. I'm really pleased to hear that you've been
>> exporting between the two programs successfully. Could you give me some
>> information on how you set up your tiers and linguistic types in the first
>> place? When I go to export to FLEx, I cannot move through the three or four
>> windows because no matter which way I go about it, the 'next' button is
>> always greyed out for me. After reading your email I tried using just two
>> tiers, with the first a parent of the second, but I still couldn't find a
>> way through to FLEx.
>>
>> Cheers,
>>
>> Jonathon
>>
>>
>>
>>
>> On 11 September 2013 22:25, Beth Bryson <Beth_Bryson at sil.org> wrote:
>>
>>> Thanks, Kate, for sharing your experience!
>>>
>>> Brighde, I have a feeling that some of the issues you are having have to
>>> do with the settings for the writing systems.  Look carefully at what you
>>> have set as your vernacular languages as compared to analysis languages.
>>>  In FLEx, it will only do analysis on vernacular languages, and in a single
>>> text, there is a single writing system that is the "baseline" writing
>>> system, and it works best to analyze that.
>>>
>>> From your description, it sounds like both Kriol and English should be
>>> analysis languages (and you wouldn't try to parse those).  It's the other
>>> language that should be the vernacular.
>>>
>>> If you want to write off-list, it's possible that I or someone else on
>>> the FieldWorks development team could help you work out the writing system
>>> settings.  And the FLEx google group is also another good place to get
>>> input like that.
>>>
>>> And everyone is right that how you set up the tiers in ELAN is crucial
>>> to this working--the interchange between the programs does make certain
>>> assumptions about that.  So it is very helpful to have Kate's description
>>> of her successful experience.
>>>
>>> Beth Bryson
>>>
>>>
>>> On Sep 11, 2013, at 12:57 AM, Kate Horrack wrote:
>>>
>>> Hi all,
>>>
>>> I'm also working with both programs, and it's all working well for me so
>>> far. In ELAN, I'm currently starting out with my tiers set up as follows:
>>>
>>> - Transcription tier: transcription of the target language (without
>>> morpheme breaks)
>>> - Blank tier (with Transcription tier as its parent): The export to
>>> flextext requires this for some reason (I've tried exporting with a single
>>> line transcription but ELAN won't allow me to do this)
>>>
>>> Then, I export to FLEx file and import this file into FLEx for
>>> morphological analysis (and the automatic parser appears to be working for
>>> me). I've also tested exporting this file from FLEx after doing
>>> morphological analysis and then importing it back into ELAN, and this all
>>> works fine too (and everything is nicely time aligned!).
>>>
>>> So, overall I'm having quite a positive experience using these programs
>>> together.
>>>
>>> If you want any further details, let me know!
>>>
>>> Cheers,
>>> Kate.
>>>
>>>
>>> On Wed, Sep 11, 2013 at 2:33 PM, Brighde Clare Collins <
>>> brighdec at student.unimelb.edu.au> wrote:
>>>
>>>> Hi all,
>>>>
>>>> I am also relatively new to ELAN and FLEX, and have recently attempted
>>>> to export my ELAN files to FLEX to use the automatic parser for my
>>>> glossing.
>>>>
>>>> Through trial and error I was able to move through the four steps
>>>> required to export the file, and created a .flextext file which was easily
>>>> imported into FLEX.
>>>>
>>>> I was able to view the exported text in FLEX, but I difficulty with the
>>>> morpheme breakdown. Usually with a text (for example, one I directly type
>>>> into FLEX in the 'baseline' area), morpheme breakdown is easy as one just
>>>> clicks the dropdown box in the morpheme row, and chooses 'Edit Morpheme
>>>> Breaks'. However, with my text imported from ELAN, this dropdown box
>>>> doesn't exist, which means that although I have a text in FLEX, I can't do
>>>> anything further with it. It is though FLEX doesn't recognise the words
>>>> that are in the text as units that can be broken down.
>>>>
>>>> I thought that part of my problem was that my tier heirarchy in ELAN is
>>>> possibly different from what should convert to FLEX: I have two speakers,
>>>> and am translating from one language into both Kriol and English. However,
>>>> I have just tried to create a simpler hierarchy with only one tier (with
>>>> one language in it), and now find that although I am able to create a
>>>> .flextext file, FLEX is unable to import it - well, it says that it has
>>>> imported it, but there are no words in the text I've created.
>>>>
>>>> So, sorry Jonathon, but I don't really have any suggestions to make,
>>>> but would be interested to hear if anyone has had these problems! And
>>>> perhaps more interested to hear if anyone has had any successful
>>>> communication between the two programs.
>>>>
>>>> Best regards,
>>>>
>>>> Brighde
>>>>
>>>>
>>>> On Wed, Sep 11, 2013 at 12:03 PM, Jonathon Lum <lum.jonathon at gmail.com>wrote:
>>>>
>>>>> Hi all,
>>>>>
>>>>> The latest versions of ELAN and FLEX are supposed to support importing
>>>>> and exporting between the two. In theory at least this should speed up the
>>>>> glossing process if it is possible to export a basic transcription with
>>>>> morpheme breakdown from ELAN to FLEX, then gloss it in FLEX using the
>>>>> automatic parser (and perhaps also add a free translation tier), before
>>>>> exporting back to ELAN for the final product, where one can view/listen to
>>>>> the original video/audio file with all the tiers. This would mean you don't
>>>>> have to gloss every single morpheme manually in ELAN, since the parser in
>>>>> FLEX is doing much of the work instead.
>>>>>
>>>>> Does anyone have any experience with this sort of workflow?
>>>>> Unfortunately I am quite new to both ELAN and FLEX, and most people seem
>>>>> unfamiliar with using the two programs together. There is a brief
>>>>> description in the manuals about how to import and export files, but
>>>>> unfortunately the windows I get look different to the screenshots in the
>>>>> manual. The problem might have something to do with the ELAN tiers template
>>>>> I am using, or maybe there's something else I'm missing here in terms of
>>>>> settings. I'm also curious as to whether a different program like ANVIL
>>>>> might be better for what I have in mind.
>>>>>
>>>>> I look forward to hearing any suggestions.
>>>>>
>>>>> Thanks and regards,
>>>>>
>>>>> Jonathon
>>>>>
>>>>
>>>>
>>>
>>>
>>> --
>>> Kate Horrack
>>> PhD Candidate
>>> Language Documentation Lab
>>> School of Languages and Linguistics
>>> University of Melbourne
>>> Phone: 0428463670
>>> Email: khorrack at student.unimelb.edu.au
>>>
>>> --
>>> Kate Horrack <khorrack at student.unimelb.edu.au>
>>> PhD Candidate
>>> Language Documentation Lab
>>> School of Languages and Linguistics
>>> University of Melbourne
>>> Phone: 0428463670
>>> <khorrack at student.unimelb.edu.au>
>>> Email: <khorrack at student.unimelb.edu.au>khorrack at student.unimelb.edu.au
>>>
>>>


-- 
Kate Horrack
PhD Candidate
Language Documentation Lab
School of Languages and Linguistics
University of Melbourne
Phone: 0428463670
Email: khorrack at student.unimelb.edu.au
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/resource-network-linguistic-diversity/attachments/20130913/19cfd367/attachment.html>