Measuring the Difficulty of Speech Recognition Tasks: A Comparison of Perplexity and METEOR

Conclusion: Our study demonstrates the effectiveness and practicality of a client-centric dialogue system for virtual counseling. By utilizing both human evaluation and automatic metrics, we can confirm that the fine-tuned model generates responses that are comparable in quality to human responses. With the increasing demand for virtual counseling services, our proposed system has the potential to provide more efficient and effective support for clients seeking mental health advice.