Wordle simulation system based on human brain thinking

The study of the human brain's thinking has been of great interest to researchers, in order to observe the way the human brain decodes words. In this paper, Wordle, an educational word-guessing game, is selected. Its commercial nature determines that the heat and reported results have high commercial value. In playing, the difficulty of each word is estimated, and the common method is to extract the static features of words and perform multiple regression, but this method lacks accuracy and interpretability. Therefore, this paper proposes a Wordle simulation system based on human brain thinking. Due to the self-motivating characteristics of the reported results, we build a Hawkes self-motivating-ARIMA model. The dynamic wordle database is constructed by imitating the logic of human problem solving. The dynamic priority is determined with TOPSIS using the extracted metrics. A distribution of results for the general case can be achieved. The validation is then performed using EERIE, and after validation, for the modified simulation model
$R2 = 0.75$
, the percentage distribution of EERIE is (0, 2, 10, 18, 36, 29, 5). Finally, the analysis for sensitivity is performed and the model is proved to be resistant to interference. This system can quickly and efficiently analyze the difficulty of the words as well as the method, which is advanced.