Font Size: a A A

Research On English Text Summarization And Machine Translation Based On Machine Learning

Posted on:2021-05-01Degree:MasterType:Thesis
Country:ChinaCandidate:G K ZhengFull Text:PDF
GTID:2428330647460167Subject:Computer technology
Abstract/Summary:PDF Full Text Request
With the development of the Internet,the widely available data is growing explosively.As human beings enter the era of big data,a large number of English information brings the problem of information redundancy to users,which makes it difficult for users to browse and filter information.At the same time,the accuracy of English to Chinese translation needs to be further improved.Therefore,it is very important to summarize the English text information and translate it into Chinese accurately,so as to help people acquire the key points and core knowledge of English articles effectively and quickly.To solve these problems,this paper designs and develops an English text summarization and translation system based on machine learning,which can help people get key information of English text quickly,accurately and efficiently.In view of the low accuracy of translation,the model optimization strategy and data enhancement strategy used in this paper can improve the accuracy of translation.That is,in terms of model parameters,adjusting the batchsize and other parameters during model training,as well as the alpha value and beamsearch value during prediction,can increase the Bleu value by about 0.7;in terms of model structure,the Bleu value of 0.8 can be increased by repeating transformer multi-layer representation fusion;in terms of data enhancement,the Bleu value of 0.5 can be increased by "back translation".At the same time,the translation system of "neural machine translation,statistical machine translation,word list" constructed in this paper can further improve the effectiveness of translation.Among them,neural machine translation uses transformer in tensor2 tensor,and statistical machine translation uses IBM model 5 in Moses.Aiming at the problem of quickly obtaining the core content of English text,this paper combines unsupervised machine learning textrank to extract text abstracts,which can extract key information of long text effectively,and translate English into Chinese more accurately through translation system,so as to help learners improve learning efficiency.This paper takes machine translation as the background,and combines the text summarization technology,which solves the problem of people's inefficient access to important information in English text,and also provides reference for the following multi field technology combination.
Keywords/Search Tags:Machine translation, Extract text summary, Transformer, Crawler
PDF Full Text Request
Related items