Research On Tourism User Portrait Model Based On Information Fusion

In recent years,with the continuous increase of user location information in all kinds of central servers,deep integration of offline user location information and online tourism scenario information can not only achieve the accuracy of tourism user profiles,reflect the multi-dimensional characteristics of users,meet the personalized needs of users,but also provide accurate services for tourism users,and further improve the market share of tourism industry.How to match labels accurately for users has become an urgent problem in this field.To solve this problem,based on the traditional user portrait model,this paper proposes an improved tourism user portrait model based on information fusion.As a kind of information with rich semantics,user location key contains a lot of multi-dimensional information such as user behavior characteristics,preferences and habits.Through the information fusion algorithm,the offline mobile users and online tourism users' location key points are fused.The fused user location key points feature more appropriate weight for offline mobile user labels.Finally,it provides accurate decision-making for offline mobile user portrait.The main work of this paper is as follows:1.Extract the location key points of online tourism users and offline mobile users.Combined with the characteristics of text information,this paper divides the location information into physical geographic location and content geographic location,respectively extract the location key points of online tourism users,and then use the density based location key point extraction algorithm to identify and extract the location key points of offline mobile users' track information,Finally,through the feature extraction of location key points,the valuable similarity tracks of the two types of users are extracted,and the similarity comparison between the two types of users is realized.2.A multi feature fusion algorithm of location key points is proposed.The main idea of this algorithm is to cluster and fuse the multi-dimensional features of the locus of location key points,and the fusion level occurs in the feature space.The track characteristics of location key points include track mean value,track length,direction and speed,and are compared with the typical applied track fusion algorithm DOM(Direct Optimization Method)?MPM(Max Point Method)?DPA(Data Partition Algorithm).The simulation results show that the algorithm in this paper has high efficiency,good stability and error convergence,but it will increase consumption in storage space,and the overall comprehensive performance is the best.3.An improved tourism user portrait model based on information fusion is proposed,which is different from the traditional user portrait model in that the key feature of location key points after fusion is added to allocate weight for offline mobile user tags,and the user accurate portrait is realized by integrating multi-dimensional features of time,location and space,and we choose scenic spot label and tourism preference label as the target label.The results show that the improved tourism user portrait model based on information fusion is feasible.