In the discipline of data mining, classification algorithm is a very important algorithm, and the data classification technology evolved from it has a far-reaching influence in the academic and social fields. K-Nearest Neighbor (KNN) classification algorithm is a typical representative of this kind of algorithm, which has the advantages of simplicity and high accuracy. However, when using KNN algorithm to process some data cases with equal weight classification features, the difference of feature values in the dataset will have a serious impact on the accuracy. In those cases, preprocessing the dataset will improve the classification performance of the algorithm, and normalization may be a good choice. In this article, we will use the KNN classification algorithm to deal with the Helen’s dating problem, which is a classic case, to explore the influence of normalization of the dataset on classification accuracy.