EARS is an intelligent e-mail responding system, which allows its users to receive timely replies to their computer problems and questions. This project targets the most common computer support requests currently handled by the computer center at The University of Georgia (UGA). The users of the system are UGA students, faculty and staff and for their benefit the system is designed to communicate with them in English. The paper describes the natural language processing component of the system, designed to work in conjunction with e-mail handling tools implemented by Michael Covington. When fully developed, EARS will benefit the computer center staff as well as its users in numerous ways. Most importantly, it will increase the quality of technical support service offered to the UGA community. The main goal of this project was to show that simple keyword and template techniques, used in the implementation of the natural language component of EARS, can perform satisfactorily in the limited technical domain of computer related questions handled by the computer center. Although nobody claims that keywords and templates are realistic models of the way a human mind processes language, they offer a quick way to extract useful information from natural language input and are adequate for many practical purposes.