Welcome to Francis Academic Press

Academic Journal of Computing & Information Science, 2023, 6(4); doi: 10.25236/AJCIS.2023.060414.

Word Data Research and Prediction Based on Wordle Game


Qi Yang, Yifan Fang, Yu Zheng

Corresponding Author:
Yifan Fang

College of Electronic Engineering and Artificial Intelligence, South China Agricultural University, Guangzhou, China


As a very popular game in recent years, Wordle contains many interesting rules. This paper hopes to get a mathematical model that can measure the difficulty of words by analyzing the words that appear in Wordle, and use this model to predict the difficulty of the word "EEIRE". First of all, we process the data of response times. The data with 1-6 times of answer is used as negative indicator, and the data with more than 6 times of answer is used as positive indicator.Then, we using the comprehensive rank-sum ratio evaluation method, the RSR values of each sample are calculated, and then sorted and divided into three difficulty levels: 1, 2 and 3, where "1" means simple and "3" means difficult. Finally, we code the words and train them with CNN. Finally, we predict the word EEIRE to be of medium difficulty.


RSR, Convolutional Neural Network

Cite This Paper

Qi Yang, Yifan Fang, Yu Zheng. Word Data Research and Prediction Based on Wordle Game. Academic Journal of Computing & Information Science (2023), Vol. 6, Issue 4: 106-109. https://doi.org/10.25236/AJCIS.2023.060414.


[1] Zhou Q., Shao Z., & Lin H. (2018). Grade evaluation of user harmonic hazards based on rank sum ratio comprehensive evaluation method. Power Capacitor & Reactive Power Compensation.

[2] Chua L., Roska T., Kozek T., & Á Zarándy. (1993). The CNN paradigm—a short tutorial.

[3] Chen T., Tong H., & Benesty M. (2016). Xgboost: extreme gradient boosting.

[4] Qi M. (2017). LightGBM: A Highly Efficient Gradient Boosting Decision Tree. Neural Information Processing Systems. Curran Associates Inc.