It would be great if the alias data you mentioned would also be published

[jexp]

Was seeing that e.g. Ramsay is in the data as Ramsay-Snow not Ramsay-Bolton.

So it would be good to have the aliases to search across them too.

[mathbeveridge]

I will get to work on cleaning up my alias file for public consumption.

[mathbeveridge]

I've been thinking about what to do for the aliases. It's not as straight forward as it seems (at first) because of disambiguation. For example, King-in-the-North could be Eddard or Robb, depending on when the title is used. Likewise, Jon could be Jon Snow or Jon Arryn. Even worse, "king" is more problematic: it could be Joffrey, Robb, Stannis, etc. I solved this problem by altering the input, say changing "king" to "king_Joffrey" and then adding the alias "king_Joffrey".

However, I believe that the public is more interested in the list of nicknames (much of which comes from A Wiki of Ice and Fire). So I will put together an alias list, with the caveat that it is not quite the same as the list that I used to parse the text. Or I could include the disambiguation aliases, with the caveat that those don't appear in the original text. But people may be interested in the terms that I included (maester, king, queen, commander, etc) in my manual disambiguation

[jexp]

I think the nickname + disambiguation aliases both are relevant, it's quite easy to check for the disambiguation aliases in the text by replacing "_" with " " and checking against the text / lowercase text.

So I think the full list would make the most sense, it could have a column though that says if it's a nickname or a disambiguation alias.