基于深度神经网络的单词预测系统设计与实现

Design and Implementation of Word Prediction System Based on Deep Neural Network

文本生成与预测是自然语言处理中一个重要的研究领域,具有广阔的应用前景,例如通过输入法或者检索框打字时预测下一个单词或者文字。然后人们的喜好和习惯不尽相同,传统的预测方法难以有很好的预测效果。而随着深度神经网络的发展与应用,利用深度神经网络模型的文本预测系统识别准确率和速度也极大地提高。本文训练并评估了最为流行的深度神经网络预测模型,并设计了一个单词预测系统,使用前后端分离技术,前端是一个可视化网页界面,后端采用多个深度学习模型,方便评估模型效果。

Text generation and prediction is an important research field in natural language processing and has broad application prospects, such as predicting the next word or text when typing through an input method or a search box. However, people’s preferences and habits are not the same, and traditional forecasting methods are difficult to have a good forecasting effect. With the development and application of deep neural networks, the recognition accuracy and speed of text prediction systems using deep neural network models have also been greatly improved. This paper trains and evaluates the most popular deep neural network prediction model, and designs a word prediction system, using front-end and back-end separation technology. The front-end is a visual web interface, and the back-end uses multiple deep learning models to facilitate the evaluation of model effects.