Deep Feature-Based Text Clustering and its Explanation

Text clustering is a critical step in text data analysis and has been extensively studied by the mining community. Most existing algorithms are based on bag-of-words model, which faces high-dimensional sparsity problems ignores structural sequence information. Deep learning-based models such as convolutional neural networks recurrent regard texts sequences but lack supervised signals explainable results. In this paper, we propose d eep xmlns:xlink="http://www.w3.org/1999/xlink">f eature-based xmlns:xlink="http://www.w3.org/1999/xlink">t ext xmlns:xlink="http://www.w3.org/1999/xlink">c lustering ( DFTC ) framework that incorporates pretrained encoders into tasks. This representations, breaks dependency supervision. The experimental results show our model outperforms classic state-of-the-art language i.e., BERT, almost all considered datasets. addition, explanation of significant for understanding principles deep learning approach. Our proposed includes an module can help users understand meaning quality