Neural Sentence Ordering with Recurrent and Convolution

作者：禅与计算机程序设计艺术

1.简介

Natural language processing (NLP) has always been an important research area in the field of artificial intelligence (AI). However, one significant problem in NLP is sentence ordering which refers to arranging the words or sentences in a natural way so that the meaning of the text is maintained and understood correctly. In this paper, we propose two neural networks for sentence ordering: recurrent neural network (RNN), convolutional neural network (CNN). These models are trained on large-scale corpora such as English Wikipedia articles, and then tested on new documents and real world scenarios. The experiments show that our RNN model outperforms traditional sequential models like HMMs by achieving better performance on various metric