A faster algorithm in French phonetic translator
As I mentioned in one of my previous posts, I wanted to change the algorithm of the French phonetic translator to make it faster. Well, now everything is up and running. The new version works 50 times faster than the old one, and it makes fewer errors too!
The new algorithm is based mainly on general French pronunciation rules (with a huge number of exceptions). The algorithm by itself is able to predict the correction transcription of French words in 96.3% of cases for regular words and 70% of cases for proper nouns. Additionally the translator uses a list of words that have such unique pronunciation that it didn't make sense to create additional rules in the algorithm (take "femme", for example). Hence, the final error percentage is much lower, at least for regular words.
Additionally I'm introducing the error correction mode. It allows you to suggest a correct transcription in case you found an error.
Another improvement – now the translator supports pronunciation variants and homographs. There's still a lot to be done with this functionality, such as an automatic detection of part of speech for better homographs conversion. I hope to implement it in the future. If you want to help me with that, please contact me.