Most of the aspell60 words get transliterated correctly. Some are transliterated incorrectly (largely acronyms or foreign words). For some others, the transliterator doesn't even give the correct transliterator as a possible option.
I need to look through the transliterations I did to find ways to improve this. This would include:
Categorizing the transliteration errors (was it an acronym, a French word, Italian word, just a random unusual word, etc.).
Look for patterns in the errors, and add new transliteration rules or refine existing ones.
In particular, Italian words with a cc are all transliterated incorrectly, and a new rule would help there.
Add options to the rules so it generates the correct transliteration as a possible output (even if it isn't the first one) for everything that isn't arbitrary input (such as acronyms).
Most of the aspell60 words get transliterated correctly. Some are transliterated incorrectly (largely acronyms or foreign words). For some others, the transliterator doesn't even give the correct transliterator as a possible option.
I need to look through the transliterations I did to find ways to improve this. This would include:
cc
are all transliterated incorrectly, and a new rule would help there.