基于LCS+LSTM的智能阅读交互系统设计与实现

Design and Implementation of Intelligent Reading Interactive System Based on LCS+LSTM

针对大数据时代日益严重的信息过载问题,提出基于LCS+LSTM的语义相似度计算模型,构建了面向多领域的智能阅读交互系统。LSTM模型包含了Embedding层、LSTM层、Dropout、BN正则化和全连接层5层,通过第六届全国数据挖掘竞赛以及百度WebQa提供的数据集进行训练。结果表明:训练取得了良好的效果,并使用Tkinter构建用户界面,实现了高效的智能阅读交互功能。

On the issue of the increasingly serious information overload in the era of big data,an intelligent reading interaction system based on LCS and LSTM-based semantic similarity calculation model was proposed,and an intelligent reading interaction system for multi-domain was built.Among them,the LSTM model includes the Embedding layer,the LSTM layer,the Dropout and BN regularization and the fully connected layer.It had been trained in the dataset provided by the 6th National Data Mining Competition of China and the WebQa dataset in Baidu.The results show good effect,and it's intelligent reading interaction function is high-efficiency in a user interface using Tkinter.