Automatic Generation of Multidimensional Labels of Educational Resources Based on Grey Clustering

With the rapid development of Internet+education and the implementation of the “Three Connections and Two Platforms Project”, online education resources are becoming more and more abundant, and society is increasingly dependent on online education resources. How to efficiently and reasonably classify resources and help users quickly access education resources is a research hotspot of online education. Therefore, an automatic generation method of multidimensional labels of educational resources based on gray clustering is proposed. The method first preprocesses educational resources, including word segmentation and de stop word processing. Feature words are extracted by combining TF-IDF and TextRank. Based on the extracted feature words, the improved KNN clustering algorithm CLKNN is used to cluster educational resources. The method of grey correlation degree is used to filter the text feature words that are clustered into the same type, and the selected feature words are used as their tag words to complete the automatic generation of multi-dimensional tags of educational resources. The results show that the accuracy P (Precision), recall rate R (Recall rate) and F-Measure of the automatic generation method of multi-dimensional labels of educational resources based on gray clustering are higher, which indicates that the research method can generate labels more comprehensively and accurately.