This research begins to refine a cognitive model of interactive information seeking for use in the design of information retrieval (IR) systems. The research reported in this paper critically examines several interactive information seeking and retrieval models that have emerged over the last two decades. Building primarily on the stratified model, interactions are identified at the surface stratum that can be utilized to design information retrieval systems that infer intentions at other strata, thereby better assisting users during the search process. This paper presents (1) the development technique utilized to modify and (2) the existing IR system that monitors these interactions and recommends search tactics to the user. The result is an increase in the performance of an IR system, as measured by precision. (Contains 28 references.) (Author)