Abstract
The decision-making process in any environment, including the library, is potentially subjective and arbitrary. The impetus for this research has been to eliminate the discretionary component in vendor selection by designing a quantitative decision-making model. Technically, the model is based on the artificial intelligence concept known as an expert system. The model uses rules provided by the expert, policies of the individual institution, and quantitative decision techniques to assign orders to vendors. Advantages to this knowledge-based model are the following: (1) assurance of quantitative objectivity; (2) efficient allocation of limited resources; (3) improved use of personnel time; (4) creation of a database of vendor price structuring; and (5) usefulness as a training resource.