View on GitHub

linked-amarna

Tools and results from parsing Shlomo Izreel's transcriptions of the Amarna Letters

Amarna Personal Names

Tools and results from parsing Shlomo Izre’el’s transcriptions of the Amarna Letters. Izre’el’s transcriptions can be found here.

Linked Amarna

Process for Data Collection and Curation

Order of Scripts:

name_greper.py. This script downloads the text from Shlomo Izre’el’s website, or uses a text file stored locally. It then prepares and generates the basic table for the database, as well as creates an Excel file that can be imported into OpenRefine. The columns might need to be adjusted in the Excel file to make them easier to work with.
OpenRefine. I ran the cluster key collision, cluster levenshtein, cluster levenshtein with the radius updated to 2, and the cluster levenshetein with block 5. I should have ran the cluster key collision in the name_greper.py script, as the OpenRefine cluster key collision is not as good as can be created in python. The following normalization function is much superior to the one found in OpenRefine.
```
 import unicodedata 

 ...

 def remove_diacritics(string):
     for c in unicodedeata.normalize('NFD', string):
         if unicodedata.category(c)[0] != 'M':
             yield c 

 ...
 ''.join(remove_diacritics(SOMESTRING)
```
After the names have been clustered in OpenRefine, export the file as csv. The current version of the OpenRefine cluster is saved as AmarnaNames4.csv.
name_database_generator.py. This function imports yaml file that I created from looking at the index of Moran’s translation of the Amarna Letters, and Hess’s Amarna personal names. The yaml file is in the data directory. This file also imports the csv file created from OpenRefine.
name_comparer.py collects text variants in the database, and offers suggestions for linking canonical forms of the name to the text line. name_insert.py was used to expand the name information, also added additional information like reference to Hess’s Amarna Personal Names.
d3_data_generator.py. This script generates the data used for the d3 visualization. The JSON file that is output from this script should be adjusted so each of the top level keys become variables, and the file needs to be renamed to graph4.js. This was for the purpose of making the file easier to import into the html page.

The Linked Amarna Visualization

The Visualization is available here.

Using the Visualization

The left hand panel contains a list of all of the names that are found in the database. Click on any of the names to display the sub graph of names connected to that person. The links between names shows that there is at least one text where both names are found. The left panel also has a search box where any form of the name can be searched, all of the transcriptions, and various spellings can be searched.

After a name is selected, a info box will appear in the bottom right hand corner. This info box can be expanded to show more information about the individual.

In the top right corner of the visualization is a toggle switch which will expand any subgraph to include friend of a friend relations. This will allow the user to see how an individuals network fits in the greater corpus.

About the Coding of the Visualization

The resulting data visualization uses a d3 force directed graph. Also fusejs is imported to allow for fuzzy searching of the names.

Bibliography

Shlomo Izre’el. The Amarna Tablets. The basis of this project was manipulating Izre’el’s transcriptions to create a database and explore it.
William L.F. Moran The Amarna Letters. Johns Hopkins University Press: Baltimore, 1992. I used the name index of Moran’s book as the primary form for each name, as well as additional information about the names (relations, and scope notes).
Richard S. Hess. Amarna Personal Names. American Schools of Oriental Research Dissertation Series 9. Eisenbrauns; Winona Lake, 1993. All names and references where checked against Hess’s book. Izre’el’s reconstruction of lacunae was kept. Also, much of the content of the scope notes goes back to Hess’s analysis. In addition, the language designation for each name is the first language that Hess mentions.
Anson F Rainey. The El-Amarna Correspondence. Edited by William Schniedewind and Zipora Cochavi-Rainey. Handbook of Oriental Studies 110. Brill: Leiden, 2015. Some of Izre’el’s transcriptions where checked against Rainey, but no new readings were adopted yet.