Research On Massive Service Identification Based On Crowd Sensing

With the rapid development of the Internet and the associated information technology,the services with the same functional properties but different non-functional properties have emerged in large numbers.The designers of service-oriented applications need to choose the suitable services from a broad pool of functionally identical or similar Web services when building applications,which means that users have more opportunities to find the services meet their needs,but also cost more time to find suitable services.Therefore,whether the service identification can efficiently and accurately meet the needs of users is crucial under mass service scenarios.However,most of the traditional service identification methods using two main service identification strategies including top-down or bottom-up service identification strategies.Traditional methods ignores the difference between the needs of users and their identification indicators.At the same time,the existing identification methods pay less attention to the feedback relationship between the user and the service.For example,users are trustworthy by default in most of the current identification approaches.As a result,services with inferior quality may be identified by mistake,which usually interfere with performance of identification results.Thus,users want to find services that truly meets their needs according to service identification approaches.In order to improve the identification accuracy,efficiency and quality,the main work of this thesis are as follows:First of all,from the perspective of users,this thesis proposes a user requirements based service identification approach for big data(URBSI-BD).In the proposed URBSI-BD,The method firstly cluster massive services with BIRCH clustering algorithm to obtain a number of service sets and then employ PSO algorithm with MapReduce mechanism to achieve a fine-grained evaluation of indicator for service identification.Finally,the method use Beth trust model on the quality of experience of users and set up a monitoring mechanism to better obtain required services.Secondly,an incentive mechanism for crowd sensing based on hybrid approach is proposed,crowd sensing tasks which are difficult to process only by computers will send to interested users.This thesis adopt a reverse auction method to construct a new task distribution mode and remuneration distribution mode,and at the same time,the quality control module will improve the perceived data quality.Finally,based on the methods and the theories above,this thesis designs and realizes the simulation system,and makes further experiments and analysis of results.The experiments results also reflect the correctness and feasibility of our approaches.