Research on passenger car recommendation based on comments mining of Internet

There are many facts need to be considered when buying a passenger car. The buyer is confused by how to make decision. In this paper, we choose car website recommendation as a decision support. First of all, various car's price, vendor, configuration, ratings and user comments and other information are collected from the car websites, and analyzed by data mining. The user's comments were analyzed to find out the user satisfaction of the various evaluation indicators of the model, extract the advantages and disadvantages of the model. And then according to the comments on the model, we revise the score to improve the accuracy by corresponding the score and comments. According to the different needs of the users, the influencing factors are divided into the basic influencing factors and the influencing factors of the mining results. Based on these factors, we establish a mathematical model, and finally obtain the recommendation function.