| （英） |
We have been developing a system to combat telephone fraud and have completed a technical feasibility study. The system analyzes conversations over telephone lines and employs natural language processing, primarily based on rules, to identify suspicious dialogue. If a suspicious conversation is detected, the system automatically calls the police or the victim's relatives on a separate line. It interrupts the conversation between the victim and the suspect, establishes a conference call, and records the incoming call information from the suspect to enable the initial investigation.
This research will be expanded to develop personal AI systems that assist individuals in their daily lives by leveraging the findings of previous studies. In today's world, often described as VUCA (volatile, uncertain, complex, and ambiguous), we face unclear guidelines for navigating our daily life. Additionally, human relationships have weakened due to the avoidance of harassment, making it difficult to receive advice, counsel, and guidance on personal behavior from others. Consequently, we are conducting research on personal AI, which can be viewed as a companion in an individual's daily life. This AI would closely monitor daily conversations and behaviors, providing spontaneous advice and support during critical situations or when inappropriate actions or statements occur. In other words, it serves as an extension of the advanced safety features found in modern automobiles, applied to everyday life. This innovative approach effectively addresses interpersonal challenges by analyzing human conversations and, when necessary, intervening to offer advice or third-party assistance, even if the user has not explicitly requested help. |