Multi-strategy KOA Algorithm for Optimizing Gated Recurrent Cell Networks in Automatic Writing Scoring Method Design

INTORDUCTION: Builds an objective, robust, high-precision automatic scoring method for essays that not only improves the efficiency of exam scoring, but also provides effective feedback to help users improve their writing skills. OBJECTIVES: Addressing the problems of current automatic writing scoring methods that fail to consider holistic and process features and lack of model accuracy. METHODS: In this paper, a methodology approach for automatic scoring of writing based on intelligent optimization algorithm to improve recurrent neural network is proposed. Firstly, relevant features are extracted by analyzing the problem and process of automatic writing scoring; then, the gated recurrent unit network is improved by multi-strategy Keplerian optimization algorithm to construct the automatic writing scoring model; finally, the effectiveness and superiority of the proposed method is verified by simulation experiment analysis. RESULTS: The results show that the scoring method proposed in this paper controls the scoring error within 0.04, which solves the problem of incomplete features and insufficient scoring accuracy of automatic scoring methods for writing. CONCLUSION: The proposed algorithm can improve the accuracy and real-time performance of automatic scoring of writing questions, but the optimization efficiency needs to be further improved.