基于深度学习的自动唐诗生成器

基于深度学习的自动唐诗生成器(论文16000字)                      
A Tang Poetry Generation Tool Based On Deep Learning   
摘 要
2016年，阿尔法围棋（AlphaGo）与围棋界顶尖高手李世石展开了围棋的人机大战，最终阿尔法围棋以4:1的得分战胜了李世石。伴随着这条新闻的传播，同时深度学习算法也便得到了更加广阔的推广。深度学习算法在许多方面都有所应用，其中比较有成熟的分别是图像识别和自然语言处理。本次毕业设计，唐诗的生成也是基于深度学习算法。通过对深度学习算法在唐诗生成方面的研究，有助于未来在自然语言分析，语言情感分析，人工智能对话等方面有更加深入的学习。以唐诗作为深度学习算法在自然语言方向上的应用，有着许多的便利之处。首先唐诗有着一定的格式以及规律，在初次学习自然语言处理时在设计模型和获取数据集方面有一定的便利之处，同时也有一定的趣味性，可以说是相当不错的设计课题。本次毕业设计主要用到的是以Python为基础的TensorFlow2.0的深度学习框架。
关键词：深度学习算法；唐诗；神经网络框架；循环神经网络

[来源：http://think58.com]

A Tang Poetry Generation Tool Based On Deep Learning
ABSTRACT
In 2016, alphago and Li Shishi, the top player in the go industry, launched a man-machine battle in go. In the end, alphago defeated Li Shishi with a score of 4:1. With the spread of this news, at the same time, deep learning algorithm has been more widely promoted. Deep learning algorithm has been applied in many aspects, among which the mature ones are image recognition and natural language processing. In this graduation project, the generation of Tang poetry is also based on deep learning algorithm. Through the research of deep learning algorithm in the generation of Tang poetry, it will help to have more in-depth learning in natural language analysis, language emotion analysis, artificial intelligence dialogue and other aspects in the future. Tang poetry as a deep learning algorithm has many advantages in the application of natural language. First of all, Tang poetry has a certain format and rules. When learning natural language processing for the first time, it has certain convenience in designing models and obtaining data sets. At the same time, it also has certain interest, which can be said to be a quite good design topic. This graduation project mainly uses the Python based tensorflow 2.0 deep learning framework. [资料来源：http://THINK58.com]
Key words：Deep Learning Algorithm；Tang Poetry；Neural Network Framework；Recurrent Neural Network

目  录
第一章 绪论 ……………………….……………………..…………………..………………1
   1.1  课题背景………………………………………..……………………………….…...1
1.2  课题研究动态………………………………..………………………………….…...2
   1.3  初步设计方法和实施方案………………..………………………………….…...…3 [资料来源：http://THINK58.com]
   1.4  本文研究内容…………………………..……………………………………………3
第二章 课题开发环境介绍……………...…………………………………………….…...…5
   2.1  Anconda简介……………...………….……………………………….………..……5
   2.2  PyCharm简介……..……….………………………………………….……..…....…5
   2.3  TensorFlow2.0简介………..………………………………………….………..……5
第三章 课题设计分析…………………………..……………………………….……………7
   3.1  课题设计可行性分析……….………………………………………….……………7
   3.2  功能需求分析……………….………………………………………….……………7
   3.3  课题设计运行环境………….………………………………………….……………7
第四章 课题程序设计…………………………..………………………………….…………9
   4.1  程序的设计思路  ………….…………………………………………….…………9
   4.2  数据集设计…………………………………………………………...….…………11
   4.3  神经网络设计………………………………………………………...………….…14
第五章 课题实现……………………………...………..………………………...…….……18 [版权所有：http://think58.com]
   5.1  程序的模块和功能  …………..……………………………………….……….…18
   5.2  唐诗生成的实现…………………..……………………………………..…………19
   5.3  程序完善……………………………..…………………………………..…………20
第六章 未来展望……………………………………….……………………………………22
   6.1  唐诗感情色彩情感分析…………………..……………………………..…………22 [资料来源：http://www.THINK58.com]
   6.2  指定题材的唐诗生成………………………..…………………………..…………22
   6.3  模型优化创新的思考…………………………..…………………………..………23
结论 ……………………………………………………….…………………………………24
参考文献 ………………………………………………….…………………………………25 [版权所有：http://think58.com]
 
 

[来源：http://think58.com]