Research On Food Online Order Based On Association Analysis And Discriminant Analysis

Nowadays,human beings have entered the Internet and Internet of Things era.The rapid development of e-commerce has brought convenience to people,and online shopping has gradually become the mainstream way.From clothing to basic food online shopping,people enjoy more convenience.More and more people choose Meituan.com,Hungry.com,Baidu glutinous rice,etc.to order food online.A lot of commentary information appears in people's field of vision.People make purchasing decisions through these comments,but many comments are long and complicated.It is not possible to see the emotional trend at a glance,and the user usually makes a purchase decision after comprehensively analyzing the emotional orientation of multiple comments.Most users who choose to take out food or food group purchases are office workers.Saving time is very important for them.Because of time,users can't read comments one by one,so it is necessary to quickly classify comments.In this paper,by using the web crawler technology,taking the comment data of the food network group purchase as an example,the machine learning based method and the semantic-based method are used to classify the emotions,and the food network group purchase dessert exclusive comment emotion dictionary is created.The classification effect shows that the semantic-based sentiment discrimination method is slightly better than the machine learning method in the case of data set imbalance;in the case of data balance,the machine learning-based classification method is better;in the machine learning-based method,random The forest discrimination method has the best classification effect.In addition,there is a certain relationship between goods and commodities.In order to solve the big data association problem,this paper selects the N merchants outsourcing order data of the online take-out platform for correlation analysis.It is mainly aimed at the association rules between dishes and dishes.It explores its relationship from different angles,establishes a dish order recommendation system,and provides the basis for the sale of food products for some non-package merchants.For many package merchants,they generally pass Experience to define the package,and the quantitative correlation analysis of this article also provides advice for this part of the business,which can be used as a supplement to its package strategy.