To simplify the capacity and processing of a program and a database constituting a device by defining three parts which are a recognition dictionary, answering operation execution, and a speech synthesis part together in an object definition table.

One object is constituted by making three parts, i.e. words and phrases constituting the recognition dictionary, operations corresponding to the respective words and phrases of the recognition dictionary, and answer-back character strings corresponding to the words and phrases of the recognition dictionary correspond to one another, and a plurality of objects are integrated to constitute a defined object definition table 31. In any of processing of a speech recognition processing part 12 specifying an inputted speech, an answering operation processing part 14 which performs operation processings of equipment to be operated corresponding to an inputted speech, and a speech synthesis part 17 for repeating the inputted speech, processings corresponding to the inputted speech are carried out by reference to the object definition table 31.