This week we continue our work on publications and proveniences. I’ve also fixed issue #1053.
The infrastructure for connecting entities and publications should be already in place. However, the biggest challenge is to actually find connections between publications and proveniences.
I tried to do some exploration on pdf mining (which inherits from AWCA). But a lot of them are in the transliteration format, so not sure what we can do with those here.
I tried to contact Rune on his idea of this method of finding “high frequency occurence”. His response - this is interesting, but we need to dig deeper into the actual meaning of the publication. Also Emilie mentioned that different assyriologists will have different words to describe a provenience.
| # | Day | Date | A short description of the work done |
|---|---|---|---|
| 1 | Monday | 2022/08/01 | Issue 1053 fix. Discussion on further work on infrastructure code. |
| 2 | Tuesday | 2022/08/02 | Exploration on pdf mining. |
| 3 | Wednesday | 2022/08/03 | Figured out that we should ditch ML-based mining. Just using regex might be better. |
| 4 | Thursday | 2022/08/04 | Ran provenience finding algorithms on existing text files of publications. |
| 5 | Friday | 2022/08/05 | Compiled an initial set of publication and provenience relationships. |
| 6 | Saturday | 2022/08/06 | Contact Rune for his opinion on this matter. |
| 7 | Sunday | 2022/08/07 | Weekend break. |