Currently, effect evaluation for teaching plan is based on certainty method, which cannot describe fuzziness and randomness in the process that students' scores indicate their knowledge grasp extent. An uncertainty-based method is proposed to mine and forecast students' knowledge grasp status. Cloud of knowledge acquisition status is designed to describe fuzziness and randomness of knowledge grasp extent. A cloud-based mining and forecasting algorithm is proposed for knowledge grasp extent. Positively producing a drop through the knowledge grasp cloud, a value indicating the effect of current teaching plan to new students, and its certainty degree, can be generated. Taking the course of principles and applications of microcomputer as an example, effectiveness of the proposed method in mining and forecasting of knowledge grasp extent, and in teaching plan reformation, is demonstrated.