Sensing Of External Service Change And Dynamic Service Recommendation

In recent years,the rapid development and application of advanced technologies such as "cloud,intelligence,things and mobile" in all walks of life have led to the trend of "every thing is service".More and more services are available in various fields,and the interconnection between services forms the service Internet.In the service Internet,service presents two characteristics,one is richness,the other is dynamic.These two characteristics also bring some problems and challenges.On the one hand,the dynamic nature of the service may cause users to be unable to understand and master the changes of the service,so that they cannot make timely adjustments according to the changes,affecting the user experience.In order to know the changes of services and inform users in time from the published information,it is necessary to study the sensing of services change.On the other hand,the richness of services leads to serious information overload problems.It is difficult for users to find services they are interested in in the massive services.Some excellent services that really meet the needs of users will also be submerged because they are not explored.In order to help users quickly and accurately find suitable services,and increase the exposure opportunities of the outstanding services,dynamic service recommendation technology for users is particularly important.Aiming at the above two problems,this paper takes mobile service and O2O service as the research object,and carries out a series of researches from the following aspects:(1)In the mobile service scenario,first of all,based on the data provided by the third-party mobile service aggregation platform,application store,the specific implementation of each sensing method is given from three aspects: the joining and exiting of mobile service,the popularity of mobile service,and the change of mobile service function.Then,in order to recommend mobile services to users individually,a model called personal service ecosystem is introduced to model users' preferences for mobile services.Finally,on the basis of bi-directional perception by using the method of sensing of mobile service change function and the model of personal service ecosystem,this paper innovatively proposes to use the personal interest evolution network model to capture the relationship between the change track of mobile service function and the user's mobile service preference track,and further proposes the move based on bi-directional sensing with personal interest evolution network as the core Dynamic service recommendation methodframework.In addition,in order to verify the effectiveness of the proposed method,a large number of experiments have been carried out,At the same time,tools have been developed to show the results.(2)In the O2O service scenario,the sensing of O2O service change is studied.However,due to the sparsity of the data related to O2O services in the actual situation and mostly only related to the service functions,only the function changes of O2O services are concerned.For the sensing of O2O service function change,this paper creatively decomposes the classification task of function change type and the recognition task of change function,and proposes a recognition and classification model based on BILSTM + CRF and TWOLAYERLSTM to solve this problem,and further evaluates the effect of the sensing method through experiments.At last,showing results with developed tools.