An Intelligent Voice Interaction System Based on Intent Recognition and Large Model Technology in the Context of Electrical Inspection

With the advancement of large language models, the importance of optimizing question-answering systems using these models has become increasingly prominent. To achieve efficient and accurate electrical equipment inspection in the context of electrical safety checks, we propose an intelligent voice interaction system based on intent recognition and large model technology. This system integrates domain knowledge such as the electrical inspection process and conventional electrical knowledge base into the dialogue system, uses the BERT-BiLSTM model to classify user intents, and then adopts different dialogue generation strategies according to different intents, thereby solving the problem that large language models cannot answer knowledge in the specific field of electrical safety inspection. Finally, through the implementation of the interaction system in the context of electrical inspection, we verified the feasibility of this method.