Research Of Expert’s Recommend Strategies In Mobile Context-aware Learning Environment

With the development of computer technology, network, mobile communication, information processing technology, teaching and learning mode has gradually developed, based on the classroom education model a variety of education model have been developed, such as educational television, remote network education, mobile education, classroom education model have the advantage of centralized, real-time, interactive, inspiring etc., which is the major teaching mode used in the world, but it can’t meet the demand of the diversity of time and place in adults education. While the development of radio and television education, network-based distance education provides a good technical and service support for adult education, but it can’t meet the demand of learning anytime and anywhere. Although mobile education can meet the demand of learning anytime and anywhere, but it can’t meet the requirements of real-time transmit of scenarios, the main contradiction is unable to meet the real-time communication bandwidth. In recent years, with3G,4G wireless broadband network application and popularization, it could meet the demands of real-time, efficient, and reliable quality of service in mobile distance education, which could better meet the needs of different learning styles, it also could further expand and enhance the student’s learning efficiency, and provide a solid material foundation for situation awareness and mobile learning. However, how to efficiently recommend appropriate number of experts to the learners is one of the key issues in mobile context-aware learning, so experts recommend strategies was came up in this paper in mobile context-aware learning.Based on the large number of literature reading, recommend strategies and learning theory were elaborated in this paper, and the expert evaluation factors in mobile context-aware learning were analyzed. Four key factors were elaborated based on the expert evaluation factors in mobile context-aware learning, the four key factors are expert knowledge proficiency TE, the average time to solve the problem of expert’s history TT, communication costs per unit time between the expert CF and learner and the expert service fee TF. With the support of expert’s real data, the four key factors and satisfaction factor were comparative analyzed, and reach the conclusion that the linear relationship exists between every two factors, then can get another conclusion:the expert’s satisfaction can be expressed as a linear combination by the4factors. Based on this conclusion, this paper proposed a multiple linear regression model of expert recommendation, and got the maximum likelihood estimator of partial regression coefficient based on actual data of the experts, finally recommend suitable experts to learner with the recommend strategy, and then record the expert service data.In order to verify the feasibility of the multiple linear regression recommendation strategy in mobile context-aware learning environment, a realistic simulation experiment was designed, and the purpose of this experiment is to guide learners to master the use of home health smart instrumentation. The experiment was arranged in two closed lab, the two labs are separated to prevent the impact of the face to face exchange of experts and learners on the experimental results, and experimental unified communication tools using the Android smart phones, all smart phones using3G wireless networks. The purpose is to guide the learner to master home health smart instrumentation’s operations, and the experimental data will be recorded.Through the analysis of real data and simulation data, here draws four major conclusions:①With promotion of expert’s knowledge proficiency TE, service fees of experts TF, expert’s satisfaction D are all rise, but expert’s services time TT was decreased.②Expert’s satisfaction D in mobile context-aware intelligent learning environment under the strategy recommendation is lower than face to face guidance, long service time than face to face guidance, but whether expert’s satisfaction or service time, it’s very close.③Expert’s satisfaction D with model predictions are very closely to actual expert’s satisfaction, this indicating that the recommendation strategy is very accuracy, learners could obtain high satisfaction specialists to provide guidance services.④This recommendation strategy could get very high expert’s satisfaction feedback, but random recommendation could get a uneven expert’s satisfaction, and very lower than recommendation strategy.