Knowledge is modeled in various data representation formats like semantic network, decision table, decision tree, etc., that attempt to capture knowledge which human perceives in his natural language. It is the depiction of knowledge that causes a lot of complexities in inferring and queering to the inference engine. Frames are the extension of semantic network in structural form. They have more expressive power in the context of inferring as they exist in object oriented form. The novel contributions of the research are proposed system architecture, proposed transformation algorithm; and system simulator design. The system performance has been tested on nine case examples; three each from the small, medium and large classes. The results were verified qualitatively by three experts and they have verified that for each case example the number and names of the nodes/frames is the identical; the number and type of attributes, methods for each node were identical; and the type of relationship between any of the two nodes were also identical in the input and the output. Hence the achieved transformation is accurate and automatic.