一种面向数据场景化使用的跨数据空间搜索系统OA

With the advancement of national policies on data factors,efficient discovery and convenient utilization of data resources across data spaces scattered in various industries and fields have become a key for the effective circulation of data elements.A data space search system oriented to scenario-based data usage was designed and implemented.In response to the security requirements that data shall not leave its domain or be aggregated in each,a distributed search framework based on the digital object architecture(DOA)was constructed.Based on the pragmatic data relationships abstracted from the scenario-based data usage,a digital object ranking algorithm,named DO-Rank was proposed.It considered both query similarity and data pragmatic relationships.It provided upper-layer applications with the capability to search and discover data resources across different data spaces.Experiments results demonstrated that without data aggregation,the system achieved performance indicators including precision,recall and normalized discounted cumulative gain(nDCG)comparable to those of centralized data retrieval methods.