Ingest Crito alignment from Ugarit

[jacobwegner]

@farnoosh-shamsian will provide the link to Ugrait (or the CSV output). As long as we are using the same Greek version from scaife.perseus.org (urn:cts:greekLit:tlg0059.tlg003.perseus-grc2), we should be able to ingest things.

[jacobwegner]

The URL is:

https://ugarit.ialigner.com/userProfile.php?userid=40&tgid=8612&p=7

We will email Tariq about setting up an export.

[jacobwegner]

@farnoosh-shamsian: I spent some time on this today.

Should have something to show soon and can follow up with Tariq about the longer term needs for an export.

[jacobwegner]

I'll start building out some metadata to go with this:

ugarit_url: "https://ugarit.ialigner.com/userProfile.php?userid=40&tgid=8612"
label: "LREC Crito"
annotator: "Farnoosh Shamsian"
scaife_reader_url: "https://scaife.perseus.org/reader/urn:cts:greekLit:tlg0059.tlg003.perseus-grc2:43-48/?highlight=@ΣΩ[1]-ἐναντίων[1]"
versions:
  - "greek"
  - "eng"

[jacobwegner]

@farnoosh-shamsian and @gregorycrane: I've done an initial pass at importing Farnoosh's annotations from Ugarit.

You can view them here:

https://bt-crito-alignment.herokuapp.com/reader/urn:cts:greekLit:tlg0059.tlg003.perseus-grc2-8612:43a?mode=alignments&rs=urn%3Acite2%3Ascaife-viewer%3Aalignment.v1%3Acrito-greek-english-word-alignment-7b34509f15734bd7a20b873aeb08eaa1

(As I mentioned earlier in this thread, I do want to work on a repeatable process for bringing these type of annotations into Beyond Translation, but I wanted to get the annotations deployed sooner than later to demonstrate we can indeed bring the annotations in!)

[jacobwegner]

(Updated the issue title because what is being ingested is a Greek / English alignment, not Persian)

[jacobwegner]

Just riffing a bit should we choose to do a deeper integration:

• We could make a small webapp available for a user to browse / select an "alignment group" on Ugarit
• We would allow the user to modify metadata about the alignment group
• Optional: Allow the user to map alignments to textparts / edit citation metadata
• The parser could then extract the alignments, create a branch on beyond-translation and open a pull request
• Further edits could be done on a fork

[jacobwegner]

@farnoosh-shamsian I was wondering if this was an alignment done by one of your students?

https://ugarit.ialigner.com/userProfile.php?userid=63464&tgid=9576

If so, would you like me to try bringing it into Beyond Translation?

[jacobwegner]

I went ahead and loaded it (to take another pass on the Ugarit tooling):

https://bt-crito-alignment.herokuapp.com/reader/urn:cts:greekLit:tlg0059.tlg003.perseus-grc2-9576:43a?mode=alignments&rs=urn%3Acite2%3Ascaife-viewer%3Aalignment.v1%3Acrito-greek-farsi-word-alignment-7b34509f15734bd7a20b873aeb08eaa1

[farnoosh-shamsian]

I went ahead and loaded it (to take another pass on the Ugarit tooling):

https://bt-crito-alignment.herokuapp.com/reader/urn:cts:greekLit:tlg0059.tlg003.perseus-grc2-9576:43a?mode=alignments&rs=urn%3Acite2%3Ascaife-viewer%3Aalignment.v1%3Acrito-greek-farsi-word-alignment-7b34509f15734bd7a20b873aeb08eaa1

This is Great! There may be some mistakes considering that both the translation and the alignment are done by a student. Eventually, we would have three Persian translations aligned with the Greek, but for now, this is fantastic! Also, I realized there are some small inaccuracies in the English alignment of Crito. I will keep track of every correction I'll make in Ugarit.

[jacobwegner]

@farnoosh-shamsian Sounds good!

I saw you had emailed some corrections. I can just update directly from Ugarit once corrections are made, so I'll pull those in and bump this thread when done.

[jacobwegner]

@farnoosh-shamsian: Loaded your latest changes:

https://bt-crito-alignment.herokuapp.com/reader/urn:cts:greekLit:tlg0059.tlg003.perseus-grc2-8612:43a?mode=alignments&rs=urn%3Acite2%3Ascaife-viewer%3Aalignment.v1%3Acrito-greek-english-word-alignment-7b34509f15734bd7a20b873aeb08eaa1

[farnoosh-shamsian]

@jacobwegner I'm aligning the final translations of Crito in Ugarit. I align them sentence by sentence and not based on the dialogues. Speaker names are included. We have three translations for each sentence. You can find them here: https://ugarit.ialigner.com/userProfile.php?userid=40&tgid=10185, https://ugarit.ialigner.com/userProfile.php?userid=40&tgid=10186 and https://ugarit.ialigner.com/userProfile.php?userid=40&tgid=10187 The title for each alignment includes the sentence number in Leuven treebanks of Crito (I don't have a link to their github but I'm working on this copy: https://docs.google.com/spreadsheets/d/1IEak-5LcvQZgcmHA4KBB5mQ_f2Geo7R54NXNnenRhi0/edit#gid=1218501027), the paragraph number in Perseus and finally the version of the translation (Tr1, Tr2 and Tr3). See also the 3rd spreadsheet here: https://docs.google.com/spreadsheets/d/1swcxA6JBgimUDtHARuPsCls5vOR2p9LO-sUBvurQfWg/edit#gid=789418587

[jacobwegner]

TODOs from our call today:

Viewing annotations in Beyond Translation

• [x] Fix sentence ordering (@jacobwegner)
• [x] Fix duplicate text part refs (@jacobwegner will split 43c.3 as 43c.3a and 43c.3b for now, at some point we should correct in Ugarit)
• [ ] Explore a four way display (grc, far[tr1,tr2,tr3]) @jacobwegner

Attributions

• [x] Provide attribution information (@farnoosh-shamsian)
• [x] Display attribution information (@jacobwegner)

Translation Commentary

• [ ] Collect "translation commentary" (@farnoosh-shamsian)
• [ ] Ingest and display "translation commentary" (@jacobwegner)

[jacobwegner]

@farnoosh-shamsian I've worked through two of the issues discussed on the call (renumbering 43c.3 and ensuring we sort the text parts by sentence number):

https://bt-crito-alignment.herokuapp.com/reader/urn:cts:greekLit:tlg0059.tlg003.perseus-grc2-10185:43a?mode=alignments&rs=urn%3Acite2%3Ascaife-viewer%3Aalignment.v1%3Acrito-greek-farsi-word-alignment-tr1-7b34509f15734bd7a20b873aeb08eaa1

[farnoosh-shamsian]

Thanks! I think I'll correct the issue with 43c.3 anyway, just for consistency, and I'll make sure the rest of them are numbered correctly.

[jacobwegner]

Sample attribution information:

https://orcid.org/0000-0003-3743-4278

[jacobwegner]

Attributions will be built here for now:

https://docs.google.com/spreadsheets/d/1_ZD7RIBCdW4gV4g_Cjk3XZIG4_hjpr-HY26l2A9Qips/edit#gid=0

Longer term, a YAML snippet or header in a Markdown file could be used to populate attribution info.

[jacobwegner]

@farnoosh-shamsian: I've loaded the attribution information you provided (and removed the "Sarabi alignment"):

https://bt-crito-alignment.herokuapp.com/reader/urn:cts:greekLit:tlg0059.tlg003.tr3-far1-10187:43a.1?mode=alignments&rs=urn%3Acite2%3Ascaife-viewer%3Aalignment.v1%3Acrito-greek-farsi-word-alignment-tr3-7b34509f15734bd7a20b873aeb08eaa1

[farnoosh-shamsian]

I have finished aligning the rest of Crito in Ugarit, in this folder: https://ugarit.ialigner.com/userProfile.php?userid=40&tgid=10320 This is in continuation to the parts that we have already ingested in Scaife (you can find them here: https://ugarit.ialigner.com/userProfile.php?userid=40&tgid=8612)

[jacobwegner]

Nice work @farnoosh-shamsian 🎉 🎊 !

I will get these on the preview instance soon and give you a chance to take a look before we deploy to https://beyond-translation.perseus.org.

[jacobwegner]

@farnoosh-shamsian: Apologies for the delay; I've updated the preview instance. When you have a moment, please take a look and let me know if things are working as expected:

https://bt-crito-alignment.herokuapp.com/reader/urn:cts:greekLit:tlg0059.tlg003.perseus-grc2-8612:50d?mode=alignments&rs=urn%3Acite2%3Ascaife-viewer%3Aalignment.v1%3Acrito-greek-english-word-alignment-7b34509f15734bd7a20b873aeb08eaa1

If they are, I can continue on with the production deploy.

Thanks!

[jacobwegner]

@farnoosh-shamsian has provided a few new links; I'll work on getting the latest translations synchronized and deployed to Beyond Translation.

Will also be a chance to update #90 for multiple languages (e.g., the Persian localization, the Portugese localization, etc)

[jacobwegner]

I've picked this up again; will have a new deploy preview that shows the "sentence-level" alignments and will work at getting other things brought forward too.

[jacobwegner]

@farnoosh-shamsian: I've deployed the "sentence-level" alignments:

https://bt-crito-alignment.herokuapp.com/reader/urn:cts:greekLit:tlg0059.tlg003.perseus-grc2b1:1-10?mode=alignments&rs=urn%3Acite2%3Ascaife-viewer%3Aalignment.v1%3Acrito-greek-farsi-sentence-alignment

I'll ping you again when I have more of the word-level data loaded.

[farnoosh-shamsian]

@jacobwegner Perfect! Thank you. All three translations are in the same spreadsheet, titled as "Primary translation", "Literal translation" and "Secondary translation". The corresponding alignments in the UD spreadsheet have the same titles.