The prediction and prevention of the initial lapse--which is defined as the first lapse after a period of abstinence--is important because the initial lapse often leads to subsequent lapses (within the same lapse episode) or relapse. The prediction of the initial lapse may allow preemptive intervention to be possible. This dissertation reports on a predictive modeling study of the initial alcohol lapse after patient left residential care. The data were collected via a mobile health application, Addiction-Comprehensive Health Enhancement Support System (A-CHESS). A-CHESS was designed to offer ongoing support to alcohol addiction patients who have returned to their own community after completing inpatient treatment programs. Patients may access A-CHESS services at any time anywhere to help them cope with the recovery. In addition to the first chapter as the introduction of the problems and context, this dissertation consists of the other three chapters; each chapter presents a study to address different challenges faced in the development of such a predictive model in A-CHESS. The first challenge is the validation of a survey instrument used in the A-CHESS Weekly Check-in; the second challenge is the study of A-CHESS use behavior before the initial lapse; and the third challenge is to develop a comprehensive predictive model of the initial lapse. The steps taken in this dissertation to address these challenges have been fruitful. The major findings are practical and can be implemented in A-CHESS. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]