Wordle word difficulty classification based on K-means

Abstract

Wordle is a popular word-guessing game, and the analysis of Wordle games will play an important role in updating its iterations. In this paper, a word difficulty classification model based on K-means cluster analysis is developed. Clustering the dataset of the number of times required for word guessing, the words can be classified into three categories according to their difficulty: hard, medium and easy, with the corresponding labels of 3, 2 and 1. The attributes of the words in each category were counted and analyzed to arrive at the following: 1) The more common the word, the less difficult it was. 2) The more repeated letters in the word, the more difficult it was. 3) The experiment substitutes the example word EERIE into the model and the difficulty classification result is medium. The profile coefficient index of the model is 0.372. The cluster analysis method applied in this paper has good training results and is suitable as technical support for Wordle game analysis.