Research On Recommend Model Of Location Service Based On Abstract State Machine

With the coming of the intelligent information era, People pay close attention to berelated to service information of eating, live and behavior.Intelligent equipment willbecome to conveniently get these services at any time and place, and these services and isclosely related to the position, such as people travel in the outside, want to know yourposition from more recent food have those, and hope to get related recommendations withyour favorite food. And as in the process of travel we can recommend a travel route thatwe are interested.Such as we share tourism route to others for providing reference totheir trip.For these needs, it has been a research focus that through the mobilecommunication network or positioning technology get mobile terminal user’s personalposition information, combine with geographical Location information system for the userto provide location-based services (Location Based Service, LBS). LBS have determinedmobile device or user geographical position for providing related location of services thatuse personalized recommendation to varying degrees, such as Facebook, Bedo, NetEaseEight Party, where to go and so on. However, existing based on location servicesrecommended system lack unified theory model, and recommendation quality need toimprove. So we mainly aimed at how to establish a location-based services recommendedmodel with theoretical basis and research users track pattern mining. Including:1.Propose the location service recommended model based on abstract state machine.The model utilizes the abstract state machine model method to describe calculating stateof recommended location services with the simple mathematical structure, and build thetransparent recommended model of location service. The model can be used as abenchmark model that can any horizontal and vertical adjustment according to the specificneeds.We use the abstract state machine model for describing the semantic construction.The model can be described at the different abstract level.2. The location services recommend model will be expanded personalization routerecommend model, and expanded the module of the model involve with data processing,clustering, calculation of users route similarity algorithm that were adopted to researchrespectively, especially we put forward new users route similarity calculation method thelongest semantic similarity calculation algorithm (LSTS), combined with dynamicprogramming ideas to avoid repeated calculation of subsequences,is mainly improves theeffectiveness and accuracy to the traditional algorithm. 3. At last，we used AsmL to check the whole route recommend model which hadinternal consistency and rationality.The AsmL test tool generate finite state machine thatprove the system reliability. Based on the personalized recommendation model we realizedthe route travel route recommendation system.The system applied the data processingmethod with calculation user similarity LSTS algorithm that we put forward. We havecompared LSTS algorithm with calculating user similarity of traditional method. Theexperimental results show that the proposed method are better than the traditional methodin efficiency and accuracy.