Research On Personalized Recommendation Of Tourism Products Online Advertising

Tourism network advertising is one of the key means to speed up the process of tourism information,and tourism personalized information service is an important way to improve the effect of online advertising.However,more and more tourism products,the accumulation of tourism programs and the surge number of online travel,not only caused online tourists will spend a long time looking for the content to meets his needs,but also lead to information overload on the tourism website,resulting in online time cost increase and the effect of website advertising recommendation decrease.In view of this,this paper puts forward the corresponding personalized advertising recommendation based on the Ebbinghaus forgetting law and Matthew effect phenomenon,aims to improve the recommended accuracy.First,on the basis of the tourists' interest changes following with Ebbinghaus forgetting law,we integrated the tourists' interest forgetting function into the collaborative filtering algorithm,give time weight on tourists' score,thus weakening the weight of the historical score and strengthen the importance of the current score.Then,we analyzed the influence of the Matthew effect on prediction of the tourists' interest in the recommendation system,and introduced the popularity of tourism products into collaborative filtering algorithm,to increase the penalty value of popular products,reduced the impact of the popularity on tourist interest similarity.Finally,according to the influence of tourists' historical score information and the implicit information of products on tourists' interest forecasting,this paper established a personalized tourism network advertisement recommendation based on forgetting function and tourism product popularization in combination with Ebbinghaus forgetting rule and Matthew effect model,and put forward three ways to improve the popularity,in-depth analysis the popularity of the recommended impact,and combined with the relevant case for the data simulation.The results show that the prediction accuracy of the model that considered the forgetting function and product similarity at the same time and improved popularity model is higher than that of the single angle optimization model.And the prediction accuracy of the model with single angle optimization is higher than that of the traditional filtering algorithm.The optimized method of advertising recommendation not only eliminates the impact of the changes of online tourists and product popularity on recommended accuracy,but also relieves the amount of information overload and reduces the time cost of guest browsing.It provided the certain methods for the website's precise advertising recommendation.And the means to broaden the tourism personalized recommendation of research ideas.