User programmability of communication network services with associative-memory neural networks

In the realization of various kinds of services by intelligent network (IN), the key issue is user programmability, where the user actually defines/executes the services on the communication network. This paper attempts to utilize the content search, generalization and robustness of the associative memory-type neural network and presents a system where the user programmability is realized by example inputs without the programming. An evaluation by simulation is executed with the case of telephone delivery service as an example. It is shown as a result that the user programmability of the communication network service can be realized even for the user with a low skill level, allowing the fuzzy access with incomplete input information.