Intelligence Evaluation System of Chinese Composition Using Optimal Convolutional Neural Network

With the rapid growth of advanced computer technology, educational technology has impressed not only on evaluation of English, but also for Chinese composition. In China, an automatic composition evaluation system has utilized for foreign language writing teaching. Writing is significant comprehensive expertise for language learners. An automatic grading of Chinese composition consists of using Natural Language Processing (NLP) as well as Artificial Intelligence (AI) to estimate as well as score compositions. In this research, an Optimal Convolutional Neural Network (CNN) is proposed for the intelligence evaluation system for the Chinese composition. In this research, the data is collected from the Chinese characters to evaluate the performance. Initially, the data is pre-processed by utilizing the stemming and stop word removal techniques. Then, Chinese word segmentation and feature extraction using Term Frequency and Inverse Document Frequency (TF-IDF) is utilized. Finally, an optimal CNN is utilized for the evaluation and scoring of the Chinese composition. The effectiveness of the proposed optimal CNN is evaluated by using an accuracy and it achieves the 79.2 % in rhetorical approaches of Chinese composition.