In the voice systems world, the term lexicon means the set of words for which a recognition engine has access to pronunciations.

在语音系统世界里，词典这个术语表示一组词，识别引擎具有对这些词的发音的访问权。

youdao

Prompts that expect a response can either provide a list of valid words that are accepted and recognized by the voice recognition system, or you can ask for input through the telephone keypad.

这个提示需要响应提供有效单词的列表，语音识别系统可以接收并识别这些单词；也可以通过电话键盘进行输入。

youdao

Instead of voice recognition for words, itshould work for just one character at a time - such as alphabets, numbers, etc.

不同于基于字母语音识别，它的工作方式是一次读一个字符 - 就像一个字母，数字，等等。

youdao

However, English is particularly tricky for speech recognition, because it has so many short words, some real homonyms, others that become homonymous when uttered by non natives (like me) whose command of phonetics is so-so.

FORBES: YouTube's Closed Caption 'Jedi': On cc Upgrades, Searchability And Why It's Personal

These implementations may still be a few years off, but speech recognition in virtual assistants (which includes understanding context, and not just words) is paving the way for these seemingly science fiction technologies to be realized.

FORBES: Guest post written by Ilya Gelfenbeyn