Natural language hardware architecture with content-addressable memory - includes specific CPU as checking unit for each verbal meaning employed in real=time dialogue simulation
摘要
The number of micro-controllers (ROM-CPU) connected in parallel between an input sound comparator and an output speech synthesiser corresponds exactly to the number of possible meanings to be identified in the input speech. When a message is entered into the appropriate ROM-CPU an association is formed with between one and four other ROM-CPUs in sequence, according to the six interpretation criteria along preprogrammed circuit connections. The two to five words spoken in each answer in a real-time sequence give a logical answer w.r.t. the content. USE - With e.g. Air Traffic Control Language in ground-to-aircraft dialogue.