Configuration method and device for voice feature database of malicious calls

Disclosed are a configuration method and device for a voice feature database of malicious calls. The method comprises: when, during a call answering process, it is detected that a user selects a malicious call option, marking a caller call as a malicious call according to a user command, and recording a certain duration of the voice of the malicious call; then sending the voice of the malicious call to a cloud server, enabling the cloud server to perform voice analysis on the voice of the malicious call, obtaining voice feature points of the malicious call, and saving the voice feature points of the malicious call in a malicious call voice feature database.