An Ensemble Method for Selective Text Clustering Based on Spectral Graph Theory

Abstract

The invention discloses a selective text clustering integration method based on spectral graph theory. The text data set adopts K-means algorithm to generate cluster members; and the spectral clustering algorithm is used to select representative members from the generated cluster members; The selected representative members are integrated by hierarchical clustering method; the integrated representative members constitute the clustering results of this paper. It solves the problem of large amount of calculation caused when the spectral clustering method is directly applied to high-dimensional, sparse, and massive text data sets. Therefore, the use of this scheme significantly reduces the calculation time of text clustering and effectively improves the performance of text clustering. accuracy. In addition, in this embodiment, the K-means algorithm is used as the base clusterer to randomly select the initial centroids, the algorithm complexity is low, and the robustness of the algorithm is improved.

Description

technical field [0001] The invention relates to the technical field of text clustering, in particular to a selective text clustering integration method based on spectral graph theory. Background technique [0002] With the continuous development of the era of big data, text information is produced in large quantities and is rich in value. How to rationally use this text information has become an opportunity and adjustment for people. As one of the important means of text data mining, text clustering is to classify a large amount of disorganized text information reasonably through similarity judgment. Since text clustering does not require a training process, it has become an effective way to organize and summarize text information. and an important means of navigation. [0003] At the same time, clustering, as an unsupervised machine learning method with a high degree of automation, has been widely used in information retrieval, automatic multi-document summarization and ot...

AI Technical Summary

Problems solved by technology

[0005] However, the above-mentioned single clustering methods can only perform text clustering in the case of a specific size, shape, and noise clusters, and their robustness is low, which leads to the problem of low final clustering accuracy.

Images