Knowledge Base Question Answering for Space Debris Queries
This work addresses the need for engineers at space agencies to efficiently access technical knowledge from knowledge bases about orbital space debris, representing an incremental improvement in domain-specific question answering.
The authors tackled the problem of answering complex natural language queries about space debris by developing a pipeline system for the European Space Agency, which generates and executes database operations from questions, leveraging out-of-domain and GPT-3 generated data to reduce overfitting with limited in-domain data.
Space agencies execute complex satellite operations that need to be supported by the technical knowledge contained in their extensive information systems. Knowledge bases (KB) are an effective way of storing and accessing such information at scale. In this work we present a system, developed for the European Space Agency (ESA), that can answer complex natural language queries, to support engineers in accessing the information contained in a KB that models the orbital space debris environment. Our system is based on a pipeline which first generates a sequence of basic database operations, called a %program sketch, from a natural language question, then specializes the sketch into a concrete query program with mentions of entities, attributes and relations, and finally executes the program against the database. This pipeline decomposition approach enables us to train the system by leveraging out-of-domain data and semi-synthetic data generated by GPT-3, thus reducing overfitting and shortcut learning even with limited amount of in-domain training data. Our code can be found at \url{https://github.com/PaulDrm/DISCOSQA}.