Design And Implementation Of WebGIS Platform For Recommending Scenic Spots Based On Collaborative Filtering

Service platform for recommending scenic spots plays an indispensable role in accelerating tourism development,promoting regional economic growth,improving tourist travel experience and other aspects.In order to fill the gaps of spots recommendation function in current mainstream tourism e-commerce business platform and improve the status of lacking personalized applications for recommending spots,this paper evaluated the recommended algorithm as well as encoded and tested the platform program through building space-time tag data model and spots recommendation model based on the microblog data,with the proposed self-learning collaborative filtering algorithm and the intersection similarity computing method as the theory support for building spots recommendation engine,and the WebGIS technology,database technology and front-end development technology as technology support for achieving platform design so as to realize the implementation and design of the service platform for recommending scenic spots in Nanjing.Specific study contents and results are as follows:(1)Within the construction of space-time tag data model,the paper elaborated the feasibility to use microblog data as the basic data for study and application from the perspective of microblog feature and described the access to obtain the microblog data;introduced the aggregation process of microblog data as well as the space-time tag data model from three aspects of scenic spots,tourists and similar spots.(2)The author,during the model construction of spots recommendation,proposed self-learning collaborative filtering algorithm based on text segmentation and tag extraction in order to improve the problem of data sparseness and new users existed in collaborative filtering;proposed the intersection similarity computing method based on feature tags in order to solve the problem that traditional similarity measure method only applies to the problem of quantitative numeric;then built respective spots recommendation model corresponding to item based,user based and self-learing based collaborative filter.(3)In the evaluation of spots recommended algorithm,the paper respectively introduced evaluation data,evaluation index and evaluation process;obtained the result that in spots recommendation based on tags,the collaborative filtering of self-learning is obviously better than that of item based and users based,through contrastive analysis on evaluation results in accurate rate,recalled rate and interestingness,improving the problem of data sparseness and new users.(4)In the design and implementation of service platform for recommending scenic spots based on WebGIS,the paper built the spots recommendation engines based on self-leaning collaborative filtering algorithm and intersection similarity computation method,adopted GeoDataBase and MongoDB to store spots spatial and attribute data,published data service through ArcGIS Server and WCF REST,invoked ArcGIS API and jQuery to realize functions,as well as designed platform user interface by using Html,CSS and Javascript,completing design and implementation of service platform for recommending scenic spots in Nanjing.