Query Optimization Based On PKLDA Model And K-dominating Skyline Algorithm

With the development of the Internet and tourism,it is imperative to develop smart tourism.Personalized tourism has become a new trend in the development of tourism industry.Although there are many intelligent tourism systems in our country,there are still many problems.The intelligence function is not clear,users can only passively accept the travel planning provided by the website;And the sights are only displayed visually,which seriously restricts the development of the tourism industry.In order to improve the user travel experience and satisfaction.Tourism should not only satisfy the traveler's interest,but also help the user to make decisions in the perspective of the user.When users want to travel,they first search for information about the attractions.In order to satisfy users,it is necessary to combine the interests of users and the information of scenic spots for comprehensive analysis.Tourism is a destination for travel,other information also needs to be considered,such as accommodation,travel route,food,which involves multi-objective optimization problem.In this paper,we construct user interest model based on the PKLDA model,and recommend the interests of the attraction for user,and the k-dominated Skyline query algorithm is used to help the user to make decisions based on the sights information,hotel prices,and distances.The research work of this paper is as follows:1)Based on the traditional LDA model,this paper introduces prior knowledge to improve and optimize the topic model.After getting the theme of the scenic spot-the word distribution,the k-means method is used to cluster,construct the user interest model,and present the scenic spots to the user.The experiment shows that the theme model(PKLDA model)can better distinguish the theme.2)Aiming at dynamically updated data sets,this paper proposes k-dominating Skyline query algorithm to solve multi-objective optimization problems.First,the data set is preprocessed by the divide and conquer method,calculate the dominating ability of the point,and the block ordered data set is constructed.The relationship between the point dominate ability,the dimension and k are used to reduce the number of nodes,and the k-dominating Skyline query algorithm on the static data set is given.When the data changes,the algorithm of inserting nodes and deleting nodes is introduced.Finally,the k-dominating Skyline query algorithm of cyclic recursion on dynamic data sets is proposed.Experiments show that our algorithm can improve compute efficiency and achieve better performance.