Traditionally, deductive databases are designed as extensions to relational databases by either integrating a logic programming language, such as PROLOG, with a conventional relational database system that provides storage persistence needed for any database system, or by integrating an expert system with a relational database system. Deductive databases take advantage of a special kind of rule recursion called linear recursion to provide inference capabilities to improve the intelligence of the database system. The simplicity of implementation of linear recursive rules, like same generation rules, is far from the difficulty and the cost of computing the results of queries based on these recursive rules. Thus, to reduce costs and improve performance of the same generation queries, many techniques were suggested. In this paper, we propose the use of materialized views to speed up the evaluation of these queries and explain how to maintain the materialized view if the underlying base relation is updated. Finally, simulations are used to compare the materialized view approach with other approaches that are used to compute the results of the same generation queries.