Research On Emotion Classification Of Online Tourism Commentary Text Based On Improved Emointional Dictionary

The wold is changing from the era of Web 1.0 to the era of Web 2.0,and the main focus of the internet is transferred from the content to the user,we could see that more and more people are getting used to to publish their views in the internet,the internet has become a convenient platform for information exchange.The high degree of the inclusiveness and convenience from the internet,which gives us the opportunity to communicate freely in the internet,so we could say that the 21st century in which we are living now is an era of communication,and the user-generated contents(UGC)has already become the tool for disseminating information in the internet,these UGC are often with important personal feelings,which could have the research value.The rise of e-commerce under the background of the internet has led to the rapid development of online tourism in China.More and more tourists get used to obtain the relevant tourist information when they are making the traveling plans,after traveling they are also glad to share their travel experience and feelings in the online traveling website.Travel network community which has tourist contribution content and free comment is generally favored by the majority of tourists,when tourists when choose the destination,the comment information has become an important basis for lots of them.There must have a certain correlation between the mass online traveling commentary information,and it has great potential value.With the development of the concept of big data,the importance of this value in recent years has become more obvious.The key to the success of online traveling e-commerce is:how to efficiently tap online travel comments information,and intelligently analyse the traveler's feedback and their comments,thereby improving tourism products and services.While the online traveling comments are rapid growing,the content of comments are becoming more and more complicated,how to obtain the useful information of online traveling comments efficiently and accurately from this huge number of comments is the problem of the research,however it is also exactly attracting many researchers to study and research it.Comments mining mainly involve emotional tendencies analysis,feature mining,subjective content identification and so on.Meanwhile,the text emotional classification technology refers using the certain methods to classify a large number of comments positively or negatively,therefore to get the useful information from comments efficiently.Nowadays,Overseas has already had many emotional analysis studies for English travel comments,which have achieved some useful results.China,the country which has the largest number of internet users in the whole world,and is the most important tourist market,the relevant online Chinese comments text has become a very important part of international research,but until now the analysis technology for Chinese UGC is still not mature,it still stays in a development stage.As the country which have largest tourism market and huge tourism potential in the world,the study of Chinese tourists' consumption preferences and interest characteristics become an important part of the current tourism studies.However,the existing online traveling research mainly focuses on the satisfaction survey and the segmentation market,and there still are some limitations in the research time and topic coverage.Therefore,this article uses the "Sukey Miner" web crawler tool to get the tourists online comments data from the domestic popular online travel website,these online traveling comments will be cleaned and pre-treated before it can be the text of the experiment,a part of them will be the training set,and another part of them will be testing set.Based on the theory of sentiment analysis,this paper uses the training set to improve the emotional dictionary and the study of the new words of the online tourism field,and construct the online travel text emotion classification model to carry on the sentiment classification to the testing set,in order to analyze the sentiment of the online travelers.