改进二进制麻雀搜索的特征选择及文本聚类

Improved binary sparrow search algorithm for feature selectionand text clustering

针对文本中存在冗余特征影响聚类精度等问题,提出一种结合蜣螂优化算法改进二进制麻雀搜索算法的特征选择及文本聚类算法。利用基于特征词权重的适应度函数完成文本特征评估,构建矢量空间模型;引入蜣螂优化算法中的圆周方向搜索机制,改进传统麻雀搜索算法中麻雀发现者位置更新策略,并融入滚动方向机制的随机游走策略提升全局搜索能力,结合转移函数对连续型麻雀位置进行更新,得到优化的二进制麻雀搜索算法,筛选出优质特征子集;选用k-means++算法完成文本聚类。通过多种基准函数及公共数据集进行验证,结果表明:所提方法能够有效降低文本特征维度,提高聚类效果。

This paper proposes a feature selection and text clustering algorithm that improves the binary sparrow search algorithm by combining the dung beetle optimization algorithm to address the issue of redundant features affecting clustering accuracy in text.The algorithm first uses the fitness function based on the weight of feature words to complete the text feature evaluation and build a vector space model;Then,the circular direction search mechanism in the dung beetle optimization algorithm is introduced to improve the position update strategy of the sparrow finder in the traditional sparrow search algorithm,and the random walk strategy with the rolling direction mechanism is integrated to improve the global search ability.Combined with the transfer function,the continuous final sparrow position is updated to obtain the optimized binary sparrow search algorithm,and high-quality feature subsets are screened out;Finally,the k-means++algorithm was selected to complete text clustering.Validated through multiple benchmark functions and public dataset experiments.The results indicate that the method proposed in this article can effectively reduce the dimension of text features and improve clustering performance.