Research On Personalized Recommendation System Of Digital Farmhouse In Mobile

With the rapid development of mobile Internet and the further progress of digitalconstruction of rural library, the majority of farmers are no longer satisfied with thecurrent reading, and more emphasis on the perception of personalized reading service.Large amount of book resources into digital rural library, easily lead to informationtrek, current academia solutions such information overload for personalizedrecommendation technology. Traditional Internet personalized recommendationtechnology is mainly based on association rules, content-based, collaborative filteringand other hybrid recommendation method, did not fully consider the user’s contextinformation, resulting in a lack of context, can not better adapt to mobile scenariospersonalized recommendations. The main difference between mobile recommendationand traditional recommendation is that users, access the network and terminal, whichmakes mobile recommendation to increase the mobility, context-aware, terminalpersonal and so on. The farmer readers contextual information and collaborativefiltering recommendation algorithm of Slope One combining, can better adapt to thedigital rural library personalized recommendation under mobile scenarios.Through researching on the development of the current situation of rural librarydigital platforms, this paper summarizes and analyzes statistical data of telephonequestionnaire and field technical service, has found some problems about Guizhoudigital rural library, to solve these problems, proposes some policy and technicalsuggestions for improvement. And these results are data support for the digital rurallibrary personalized recommendation under mobile scenarios. Based on the aboveresearch and learn from the results of relevant research scholars, give the contextdescribe of the digital rural library farmer readers, using equivalent dissimilaritymatrix clustering method analysis user context, in order to achieve the static usercontext clustering, which is the first stage. Prediction book score combined with SlopeOne algorithm, recommend a high score book to users, eventually give a TOP-10listof recommended books, which is the second stage. To take this two-stage concertedrecommend strategy is more suitable for the digital rural library personalizedrecommendation under mobile scenarios.Based on the recommended strategy, the paper has been designed systemarchitecture and the topological structure about the digital rural library personalizedrecommendation under mobile scenarios, use UML system modeling, has designed database, achieved the system prototype, and given the corresponding evaluation.Results of this study are based on relevant scholars of researching on personalizedrecommendations. Extract static user context two-stage concerted recommend strategy,and apply the ideas for personalized recommendation system. But, the system is shortof user data and book resources, lack of researching on key technologies for thedigital rural library. In the future, do more depth discussion. Results of this study bothon theory and practice, has provided some ideas for digital construction of rurallibrary, and thinking the digital rural library personalized service.