Space Mission Data Provenance Traceability

Abstract

First, this paper proposes the concept of "space mission data provenance traceability" based on relevant concepts from other disciplines. Also the unique features of space data provenience traceability are presented. In particular, its significance to space mission interoperability is considered. Next, this paper analyzes some typical requirements for space mission data provenance traceability from different parties, including mission managers, data producers and data consumers, each of whom will benefit from space mission data provenance traceability. Then, this paper summarizes the design constraints of space mission data provenance traceability. As to the key elements for data provenance traceability, this paper refers to the "W7" model that has been accepted by many other disciplines, and analyzes its applicability to space mission data. Finally, this paper takes telemetry data as an example to identify the key problems of space mission data provenance traceability, including form of data annotations, frequency of data annotations, order and time sequence of data annotations, semantics collision avoidance of data annotations and last step of space mission data provenance tracing. The significance of this paper is to propose the concept of "space mission data provenance traceability", and to analyze its significance to space mission interoperability, which enriches the content of space data management. This paper also discusses typical requirements on space data provenance traceability, summarizes the design constraints, and identifies key problems. It is expected to provide a theoretical reference to the establishment and implementation of industry standards, and to promote the popularization and application of future space mission data.

Establishing Data Provenance for Responsible Artificial Intelligence Systems

“…Given the opaque nature of many AI-based systems, data provenance is essential for understanding AI-based systems' recommendations [74]. Recent technological advances include explainable artificial intelligence (XAI) methods, GANs, and deep learning with advances in small data techniques.…”