I'm looking for a matrix of language pairs that shows some measure of their similarity. My main issue is that I don't know of heuristics that work well across different writing styles.
Wikipedia's lexical similarity article contains a table, but it only contains languages in the Indo-European family, is sparse, and I'm not sure what the lexical similarity coefficient is based on. I found this table of average Levenshtein distances of word pairs between languages, but that only words for languages that use similar alphabets.
Specifically, I'm looking for similarities between the following languages:
- Arabic
- Bengali
- Czech
- Dutch
- French
- German
- Greek
- Hindi
- Hungarian
- Indonesian
- Italian
- Japanese
- Korean
- Mandarin
- Polish
- Portuguese
- Romanian
- Russian
- Spanish
- Swahili
- Turkish
- Vietnamese
Does such a matrix exist? If not, is there an existing dataset can I use to generate one?
