English translation of classics is an important way for Chinese culture to “go global”，Reader bet365 sign up offer are the final criterion for a translation to be accepted。The current research trend in mining readers’ opinions of English translation of classics is: under the guidance of reader acceptance theory，Introducing natural language processing technology and text mining technology，Expand the horizons of reader perspective research，Acquire large-scale comment corpus through the Internet environment，Automatic mining and quantitative analysis of bet365 sign up offer and opinions，Combining the key topics of readers’ bet365 sign up offer，So that the system can deeply discover valuable viewpoint information，Provide accurate and reliable data analysis to grasp reader bet365 sign up offer。Comprehensive use of natural language processing technology、Interdisciplinary research on machine learning methods and semantic analysis，Is an effective way to explore opinions on English translation of classics，The specific implementation path is as follows。

　　First，Constructing a corpus of English bet365 sign up offer reviews of classics，Expand the research horizons of readers’ opinions。Taking the comment texts of American and British readers as the main research object，Collect review corpus through Amazon overseas website API interface，There is currently no standard experimental corpus for English translation book reviews of classics，Use natural language processing technology to remove stop words and noise data、Preprocessing such as part-of-speech tagging，Syntactic bet365 sign up offer of sentences in context、Reference resolution and omission recovery，Establishing a corpus of English translation reviews of classics。

　　Second，Extract the thesaurus and modifiers of English translation reviews of classics，Combining the key points of readers’ bet365 sign up offer。Short text for online bet365 sign up offer、Characteristics such as sparse features，Extract the collocation relationship between subject words and opinion words through techniques such as dependency syntax analysis and semantic analysis，Explore the fixed combination patterns of subject words and modifiers，Analyze the comment topics implicit in the comment text，Recognizing text patterns of review corpus，Automatically extract explicit keywords in bet365 sign up offer，Like the translation style、structure、Language style, etc.，Explore the implicit subject words that are not directly described in the review corpus but can be derived through semantic inference。

Semantic analysis and domain knowledge representation are the keys to improving the accuracy of opinion mining in online bet365 sign up offer。The basis of semantic analysis is lexical representation，Using word vector representation in the emotional word polarity classification task、Deep learning algorithms such as feedforward neural network and convolutional neural network，Effectively improves the polarity analysis of emotional words、Accuracy rate for tasks such as semantic analysis。

Introducing the domain knowledge base to analyze the context can understand the emotions that readers really want to express，The basic task of improving the domain knowledge base is knowledge graph completion，The existing knowledge graph completion algorithm takes a long time、Limited accuracy，Interdisciplinary deep learning algorithm is an effective bet365 sign up offer way to solve this problem。

　　Third，Distinguish the emotional polarity of opinions，Creating a summary of bet365 sign up offer and opinions on English translation of classics。Comment Emotion polarity discrimination is a key step in exploring potential opinions and attitudes。From a machine learning perspective，Emotional polarity recognition can be regarded as multi-category、Single label text classification task。Machine learning classification algorithm combined with emotional dictionary，Emotional polarity that can effectively mark opinion modifiers，Provide objective data for quantitative research on the positive and negative bet365 sign up offer of readers of English translations of classics；Combined with the clustering algorithm, the intrinsic connections and objective rules between review topics can be discovered；Modifiers that express the reader’s point of view are analyzed through syntactic analysis and pattern mining at the grammatical level，Able to accurately analyze the topic summary and emotional polarity of bet365 sign up offer；Opinion topic summary based on machine learning and emotional lexicon，Exploring the underlying opinions and attitudes contained in online bet365 sign up offer，It can help translators and publishers accurately grasp readers’ positive and negative bet365 sign up offer on the translation based on credible data。

　　Fourth，Deeply dig into the semantic topics of bet365 sign up offer，Get the implicit reader’s perspective。Discovering the opinions of reviews on English translation of classics must not only focus on the style of the translation、Inspection of micro-level aspects such as word choice and sentence selection，It is also necessary to grasp the internal connection and importance ranking of viewpoints and themes from an overall perspective。Internet bet365 sign up offer are noisy、Free expression、Large scale of corpus、Opinions are sparse and scattered。To sort out the key points that readers pay attention to，Need to build an opinion topic model for deep semantic mining，Revealing review topics at a semantic level，Mapping high-dimensional review text to low-dimensional topic space，Make it more interpretable，Discover hidden valuable themes from multi-dimensional analysis，Combined domain knowledge，Classify the extracted subject words，Drawing a topic word co-word clustering map，Using visual similarity mapping technology and weighted module parameterized clustering algorithm to present topic clusters that overseas readers are highly concerned about，Present the key keywords in each cluster based on the centrality of knowledge network nodes，It can break through the limitations of original book review topic extraction limited by subjective analysis and small sample data，Extracting sentences from complex comment information has wider coverage、Implicit knowledge with richer subject vocabulary diversity。

　　Fifth，Text visual analysis，Systematic analysis of readers’ bet365 sign up offer and opinions。Integrate explicit ideas in idea summaries and implicit ideas in topic models，Based on semantic equivalence、Levels and related relationships，Merge subject words、Description and representation of superior, inferior, or related relationships；Sort the subject words in order of importance；Collect readers’ opinions on which translators、Which subject words in the translation are compared；Comparative study on reader evaluations of currently widely accepted translations of Chinese cultural classics based on thematic clustering perspective，Discover the deep-seated reasons why English translations of classics are so popular；Analyze the distribution of emotional polarity of subject words and modifiers，Understand the specific attitudes of foreign readers towards a specific translator or version，Provide scientific and reliable basis for translators and publishers to further understand readers’ needs。The principle of statistical floating can be further used to display frequently used subject words in the English translation reviews of classics in the form of a topic word cloud，And sort the summarized subject words in order of importance。Analyze the semantic relationship between viewpoint topics，Semantic description of explicit contrastive relationships in online bet365 sign up offer，Calculate the similarity between topic clusters，Semantic clustering of evaluation opinions based on deep linguistic analysis，Systematic analysis of readers’ bet365 sign up offer and opinions。

　　Sixth，Adapt to multi-bet365 sign up offer and cross-domain environment，Meet the challenges of internationalization。The international nature of the Internet determines the importance of multilingual communication、It is particularly important to study English translation reviews of classics in a cross-field context，Syntactic analysis、Basic analysis methods such as emotional polarity discrimination are highly relevant to the field of language environment problems，The emotional characteristics of data in different fields are not exactly the same，Emotion prediction model trained on data in a certain field，Usually cannot be used directly in other fields。As the number of user bet365 sign up offer and the number of various fields continues to increase，Training models individually for all domains consumes a lot of time and resources。

Cross-domain sentiment classification improves the target domain through knowledge of related source domains，Specific implementation through transfer learning or domain adaptation model in similar fields，For example, the sentiment classifier obtained by using labeled bet365 sign up offer in the book review field，Migrate or adapt to the field of digital video discs，Save time and resources on annotating bet365 sign up offer in this field。Comment sentiment usually has feature drift problems in different fields，In the field of books, words such as "readable" and "thoughtful" are commonly used to express positive emotions，Use "bland", "no plot", etc. to express negative emotions；In the field of digital video discs，Usually using "high clarity", "smooth", etc. to express positive emotions，Use “blurred”, “scratched”, etc. to express negative emotions。Due to differences between fields，Sentiment classification model trained in the source domain，Tends to perform poorly when applied directly to the target domain。Using methods based on deep learning，Can provide solutions to the problem of emotional feature drift in cross-domain environments，The difficulty that needs to be solved is how to process semantically rich comment short texts。

Cross-language sentiment bet365 sign up offer is to use source language text to analyze the emotional tendency of target language text，The specific implementation can be based on resource migration and federated learning methods。The resource migration method is difficult to implement due to different language corpus annotation systems，The joint learning method mainly relies on machine translation，Greatly affected by the quality of translation results。In recent years，Deep learning has become a hot topic in cross-language sentiment bet365 sign up offer research，Currently focusing on the coarse-grained level，Cross-language fine-grained sentiment bet365 sign up offer requires further research。

(The author is the person in charge of the National Social Science Fund Project “Research on Mining Online bet365 sign up offer and Opinions of Foreign Readers on English Translation of Classical Books”、Professor of Dalian International Studies University)