The design of teaching materials, under the construction of one outline and multiple textbook edition, has diversified the teaching materials since the implementation of educational reform. The design of various levels of contents generates a gap between academic years and textbook versions. This research, conducted by way of artificial neural network, aims at evaluating the amount and level of words of Chinese essays in elementary schools. The articles chosen from each grade are used to train a neural network that can classify the essays to the proper academic years. The proposed model can be applied to modify and analyze articles, and display its results immediately. The users can use the outcome of this research for reference when they plan to write articles for primary schools in the future.