Research On Hot Spots And Tourist Route Recommendation Based On Spatial Data Mining

With the development of science and technology and the improvement of people's living standards, travel has become a vital part of people's lives, although now there are lots of travel applications, but most of them are helpless, the search result through the search engine or travel site is either over publicized or the ads of some travel company.How to help users to find fun attractions and planning a good route to play at a strange city in an easy way? This paper investigates the current research background and the current research situation of spatial data mining and tourism recommendation at home and abroad, summarizes the related research techniques and research methods, and puts forward a new method based on spatial data mining to recommend popular attractions and route planning methods for the users to solve their problems.As many travelers have the habit of taking pictures and like to share these photos in the social media, this article excavated a large number of space with geographic information and text tag from a large-scale of geo-pictures shared at Flickr,The spatial map of the space-carrying geographic information uploaded by these users is clustered by the space where the pictures are located by using the P-DBSCAN(A Parallel of Density-Based Spatial Clustering of Applications with Noise) clustering algorithm, and then the attractions obtained from these clusters are mapped to the real attractions according to the text tag, and the scenic spots are established, and then the tourists are calculated Attractions to explore the number of hot spots, and based on the travel history and background information such as the use of collaborative filtering algorithm to calculate the user similarity,and as a basis for personalized attractions recommended in the recommended consideration of the weather and time factor. The analysis of the travel history of a single tourist can be used to extract the tourist routes,using the sequence pattern mining algorithm can get a popular travel route, and then according to the travel history and the current search text to personalize the line recommended.In this paper, the P-DBSCAN algorithm is adopted in the process of clustering. The experimental results show that the P-DBSCAN algorithm is a parallel algorithm of the DBSCAN algorithm in the case of uneven density distribution. And in the recommended process cited Apache Mahout provides a collaborative filtering algorithm recommended engine Taste,and Android devices to show the final recommendation results.