Crowdsourcing Spatio-Temporal Data Model Considering Reputation

Crowdsourcing data are contributed by non-professionals incorporating new properties such as the contributor's reputation, ans degree of trust in the contributed geographic objects. Furthermore, a crowdsourced geographic object usually has multiple versions when it is modified by several volunteers, so a mechanism for evaluating the reputation of a contributor is an alternative way to select the most creditable version. These new issues cannot be expressed and processed in traditional spatio-temporal data models. Therefore, a new crowdsourcing spatio-temporal data model is proposed in this paper, which takes reputation into consideration. The main elements in crowdsourcing data, e.g., geographic object, object status, object version, contributor, reputation, and evenst that change the state of an entity in the real world or the object in information system and their interaction mechanisms were analyzed. An object-oriented approach was used to design a crowdsourcing data model, and a UML diagram of this crowdsourcing spatio-temporal data model considering degree of trust is presented. The reputation related operations and their linkage relationships were analyzed, and eight reputation linkage operation rules were established. A prototype system using this crowdsourcing spatio-temporal data model was developed to verify the effectiveness of the model.