Because many street names are difficult to transcribe correctly after the fact—for instance, "Monsey Blvd" often gets mistakenly transcribed as "Money Boulevard" by speech-to-text —it's crucial that the system recognizes the correct names during transcription. The current models may not correct "Money" to "Monsey" since "Money" could be the intended word. Therefore, I want to input all possible street names beforehand, ensuring the transcription process selects from these pre-defined options only.