Research On Travel Recommendation Algorithms In Multiple Data Source Environments

"Intelligent tourism" accompanied by the development of Internet technology,has played a crucial role in coordinating user needs and the allocation of tourism resources.The application of the recommendation system has brought new solutions to smart travel.However,with the explosive growth of tourism data on the Internet,traditional tourism recommendation algorithms face many challenges,such as poor accuracy of recommendation and low degree of personalization.In view of the above problems,this paper studies the application of collaborative filtering technology in the recommendation algorithm in the field of tourism,and the tourism recommendation algorithm in the context of multiple data sources.The main work is as follows:In view of the above problems,this paper combines the multi-source fusion technology focused on the study of multi-source constraints under the recommendation of tourist routes,as well as multi-data source environment tourism recommendation algorithm;assisted data sources to help target data source recommendation strategy Increase recommended accuracy and personalization.The main work is as follows:1.The paper summarizes the development of recommendation system,combs the contribution of domestic and foreign experts to the recommendation system,introduces the mainstream recommendation algorithm and its implementation,discusses the advantages and disadvantages of the relevant algorithms,introduces the evaluation index of the related recommendation system,Problems to be solved.2.The problem of personalized tourist attraction recommendation service was studied.According to the data sparsity problem in the current tourism recommendation system,a collaborative filtering algorithm was proposed to integrate the features of the attraction and the attraction score.The algorithm improved the similarity calculation method in the recommendation process,combined with the features and attractions of the attraction.The scores give a global similarity,which relieves the data sparsity to a certain extent and improves the accuracy of the recommendation system.A tourism recommendation algorithm based on multiple data sources is proposed.3.Aiming at the problems of data sparsity and cold start in traditional single data environment,a multi-data source tourism recommendation algorithm is proposed based on the idea of improving tensor decomposition.The algorithm uses the auxiliary data source Evaluating the global similarity of target data sources alleviates the problem of low accuracy of recommendation due to lack of ratings for target data sources.The algorithms proposed in this paper are verified on real datasets.According to the personalized travel route algorithm proposed in this paper,the experimental results show that the algorithm can get better results in terms of driving distance and travel time.In the area of multiple data sources,the proposed algorithm effectively mitigates the problems of low recommendation accuracy and low personalization due to the lack of scoring of target data sources.