A virtual operator technique for enhancement of computer-to-computer interactivity [thesis]

Ying-Chan Fred Wu
PAGE In order to make an additional shell to deal with the problem of intersystem communication, an interpreter may be used. For each different system connected, such an interpreter may be installed to translate its commands into equivalent commands compatible to the destined system (Figure 5). Using such command translation software is a traditional way to handle the problem of system compatibility. Since the translation and service routines are fixed by the software, it is virtually
more » ... for a user to add or update the translation routine. Moreover, in the case where a new system is to be added, new translation software will be required. This suggests the concept of a "Virtual Operator" (Figure 6), an expert system software implementation based on Artificial Intelligence (AI) strategies. The research in this dissertation proposes such a method to enhance interactivity between systems. Achievement of system adaptivity to environmental changes is attained by formulating human-like learning and decision capability in the knowledge base of the virtual operator. The virtual operator will connect the terminal to the requested system, translate the user's commands to equivalent operations in the connected system, and let a user with only limited computer background communicate with 3 different systems as desired. Additionally, because of the knowledge base contained in the virtual operator, it can be trained by a human expert; experiences during the communication with different systems can be learned and accumulated to improve performance in subsequent communications. Based on the principles of building an expert system, decision rules (which control the whole learning and communication process), and application program modules are implemented separately, making it relatively easier to change the software in the virtual operator to accept a new system or a change in the command set. In order to verify the methodology of the virtual operator approach, several systems were chosen for testing, and the virtual operator now has been trained and is capable of communicating with NAS9160/WYLBUR, VAX/VMS and VAX/UNIX. r> FIGURE 1. User's view in accessing a resource Human expert Knowledge Engineer Expert System answers, solutions J FIGURE 7. Process of Knowledge Engineering
doi:10.31274/rtd-180813-9340 fatcat:u325eopjdzbwzbjrlgj37bejyu