Research On Itinerary Recommendation For User Groups In Temporary Social Network

The concept of temporary social networks has been used in hotels,concerts,sports meets and other scenes.People come to the same place and participate in the same event in a short time due to common interests or purposes.Users use their social network accounts to join temporary social networks in certain popular locations to form a social group.Existing research defines a wealth of application scenarios and constraints for single-user itinerary planning,and the problem definition for group itinerary planning is not complete.In recent years,car rental travel has become very common.A car can usually contain at least 4 passengers.Charter travel with friends can increase user satisfaction and help these users save travel time and travel costs.The degree of congestion when arriving at the POI during the trip will affect the user experience.In the existing itinerary planning system,each POI is regarded as a vertex.In fact,a POI with a large coverage area(such as Hangzhou West Lake and Hangzhou Paradise)is recommended to the user as a POI but the sub-POIs(such as West Lake:Broken Bridge,Su Causeway,Yue Fei Temple)and the travel order are ignored when the user arrives near this POI.But if the itinerary recommended to the user at the beginning is detailed to include each store and sub-scenic spot,it not only wastes the calculation cost of the recommendation system but also puts pressure on the user with overly restrictive travel constraints,because during the actual travel,the travel time may change due to uncertain factors.Therefore,this paper proposes a user grouping and itinerary recommendation method for user groups in temporary social network to help users to form a group with users with similar hobbies and free time in the temporary social network.By predicting the degree of POI congestion upon arrival,the itinerary to avoid congestion is recommended to the group.In order to provide users with comprehensive POI,accurate routes,and interactive,flexible travel services,this article proposes an itinerary-based POI recommendation and sub-itinerary planning method to predict the possible sub-POI for the user in the future,recommend a variety of surrounding attractions(shops,restaurants,etc.)for the user in the itinerary,and embed sub-itineraries for the overall itinerary to refine the itinerary.Finally,the experimental results based on real social network datasets crawled from Foursquare and Flickr show that user grouping and itinerary recommendation method for user groups in temporary social network can recommend high-quality charter groups and trips for users.Itinerary-based POI recommendation and sub-itinerary planning method can effectively provide diversified POI pushes and provide users with more accurate trip planning while ensuring recommendation relevance.