Asian languages are a lot more difficult for automated voice recognition than western languages. This is mostly due to inflections changing the meaning of the word and not just the emphasis. IBM attempted a Japanese voice recognition system in the 80's with disastrous and very humorous results. Good luck!
This isn't the 80s anymore, modern ASRs like Nuance Recognizer (which Siri uses) is light years ahead of what you experienced in the 80s. 95%+ accuracy rates!