Domain-Adaptive Fine-Tuning of Whisper for Search Query Transcription in Television Voice Interfaces