聊天机器人
原理： 严谨的说叫 ”基于深度学习的开放域生成对话模型“，框架为Keras（Tensorflow的高层包装），方案为主流的RNN（循环神经网络）的变种LSTM（长短期记忆网络）+seq2seq（序列到序列模型），外加算法Attention Mechanism（注意力机制），分词工具为jieba，UI为Tkinter，基于”青云“语料（10万+闲聊对话）训练。
运行环境：python3.6以上，Tensorflow，pandas，numpy，jieba。(Chat Robot　　　　　　Principle: Strictly speaking, it is called "Open Domain Generation Dialogue Model Based on Deep Learning". The framework is Keras (High-level Packaging of Tensorflow). The scheme is LSTM (Long-term and Short-term Memory Network)+seq2seq (Sequence to Sequence Model), plus Attention Mechanism (Attention Mechanism). The word segmentation tool is Jieba and the UI is Tkinter. Based on "Qingyun" corpus (100,000 + chat dialogue) training.　　　　　　Running environment: Python 3.6 or more, Tensorflow, pandas, numpy, jieba.)