Research Of Emotional Analysis Based On Online Reviews By Meituan Take Away Users

Due to the rapid development of 020,more and more people choose to order meals online and the review becomes the main way for users to post information after consumption.Meanwhile,users decide whether to place an order by browsing online comments.Therefore,it has become an important research topic to analyze the emotional tendency of users' online comments.This paper takes Meituan take-out APP as the object of research and analyzes the tendency of users' online comment text through data mining.Then the results of emotional inclination were studied deeply so we can put forward reasonable suggestions for merchants,consumers and Meituan.The specific research works are as follows:(1)Using Charles captures comments data.Firstly,Charles was used to fetch the users' online comment in the Meituan take-out APP.Secondly,the hash algorithm is used to carry out de-weight and automatically repair of the data.Lastly the data is saved in the nonrelational database-Mongo DB.Mongo DB reduces data processing time.(2)Using Baidu AI analyses comment text of emotion.The database was connected to Baidu AI for emotional inclination analysis.SPSS was to analyze the relationship between emotional inclination and scoring evaluation.By comparing the cloud map of negative emotion keywords,we know the users' comment emotional tendency.The evaluation of Meituan take-out aren't exactly the same and the main reason for the users' poor comments is the poor quality and quantity of dishes.(3)Contrasting users order emotion.Firstly,the thesis analyses daily order time distribution,weekly order time distribution and the change of peak order time in a week.Then it analyses the delivery time and order time in different cities.Lastly,comparing the emotional tendency of different delivery time and order time in different cities.We know the emotional value of user comments is not related to the time of ordering,but to the time of delivery.