Hybrid dimension reduction by integrating feature selection with feature extraction method for text clustering

作者：

Highlights：

• A novel hybrid dimension reduction method is proposed.

• It obtains a highly informed and much reduced feature subset.

• It improves obtained results of the underlying clustering method.

• It improves computational complexity of the underlying clustering method.

摘要

•A novel hybrid dimension reduction method is proposed.•It obtains a highly informed and much reduced feature subset.•It improves obtained results of the underlying clustering method.•It improves computational complexity of the underlying clustering method.

论文关键词：Text clustering,Feature selection,Feature extraction,Term variance,Document frequency,Principal component analysis

论文评审过程：Available online 27 November 2014.

论文官网地址：https://doi.org/10.1016/j.eswa.2014.11.038