Assalamu alaikum,
I maintain Muttaqi, an Islamic companion app on Google Play. You kindly shared
the narrator database with me previously, which is now used in the app to show
verified narrator information, so thank you again for that.
I am writing about two collections I have not been able to source reliably:
"Musnad Ahmad ibn Hanbal" and "Sunan ad-Darimi".
At the moment I use the Arabic-only CSVs from mhashim6/Open-Hadith-Data. They
work, but they have two limitations:
- **Arabic only:**There are no English or Urdu translations, so users who
cannot read Arabic get very little from these two collections compared to
the others in the app.
- No book or chapter structure: The CSVs are a flat list of hadith with no
kitab/bab partitioning, so I currently synthesise a single book spanning the
whole collection, which makes them hard to navigate.
My question:
- Do you have structured data for these two collections (Arabic text, book and
chapter divisions, and hadith numbering consistent with sunnah.com) that you
would be willing to share?
On translations, if they would be useful to you. I may be able to help, and
I want to be completely transparent about how, because provenance matters more
than the offer itself.
The route I have available is AI-assisted: a first translation pass by one
model, then a second pass by a different model checking the first, and then
review by your team. I want to be honest that I do not consider the second AI
pass to be verification. The two models share training data and failure modes,
so they tend to agree with each other rather than catch each other. Your team's
review would be the actual check, not a rubber stamp, and I would not want it
presented to users as anything other than that.
Because that puts the real work on your side, I would rather not hand you a
27,000-hadith dump. If there is interest, I could prepare a small sample,
perhaps 100 hadith with the Arabic alongside, so your team can judge in an hour
whether it is anywhere near your standard. If it is not, nothing is lost and I
will not pursue it.
I also suspect you may already have work underway on these two collections, in
which case please ignore the above entirely, and I would rather wait for
something done properly than rush something in.
JazakAllahu khairan,
Yaseen Akram
Muttaqi
support@muttaqi.app
Assalamu alaikum,
I maintain Muttaqi, an Islamic companion app on Google Play. You kindly shared
the narrator database with me previously, which is now used in the app to show
verified narrator information, so thank you again for that.
I am writing about two collections I have not been able to source reliably:
"Musnad Ahmad ibn Hanbal" and "Sunan ad-Darimi".
At the moment I use the Arabic-only CSVs from mhashim6/Open-Hadith-Data. They
work, but they have two limitations:
cannot read Arabic get very little from these two collections compared to
the others in the app.
kitab/bab partitioning, so I currently synthesise a single book spanning the
whole collection, which makes them hard to navigate.
My question:
chapter divisions, and hadith numbering consistent with sunnah.com) that you
would be willing to share?
On translations, if they would be useful to you. I may be able to help, and
I want to be completely transparent about how, because provenance matters more
than the offer itself.
The route I have available is AI-assisted: a first translation pass by one
model, then a second pass by a different model checking the first, and then
review by your team. I want to be honest that I do not consider the second AI
pass to be verification. The two models share training data and failure modes,
so they tend to agree with each other rather than catch each other. Your team's
review would be the actual check, not a rubber stamp, and I would not want it
presented to users as anything other than that.
Because that puts the real work on your side, I would rather not hand you a
27,000-hadith dump. If there is interest, I could prepare a small sample,
perhaps 100 hadith with the Arabic alongside, so your team can judge in an hour
whether it is anywhere near your standard. If it is not, nothing is lost and I
will not pursue it.
I also suspect you may already have work underway on these two collections, in
which case please ignore the above entirely, and I would rather wait for
something done properly than rush something in.
JazakAllahu khairan,
Yaseen Akram
Muttaqi
support@muttaqi.app