Combining Axiom Injection and Knowledge Base Completion for Efficient Natural Language Inference
This work addresses efficiency and scalability issues for researchers and practitioners using logic-based natural language inference systems, though it is incremental as it builds on existing methods.
The authors tackled the trade-off between knowledge data size and efficiency in logic-based Recognizing Textual Entailment (RTE) systems by replacing a search-based axiom injection mechanism with Knowledge Base Completion (KBC), significantly reducing processing time while maintaining or improving RTE performance with added knowledge data.
In logic-based approaches to reasoning tasks such as Recognizing Textual Entailment (RTE), it is important for a system to have a large amount of knowledge data. However, there is a tradeoff between adding more knowledge data for improved RTE performance and maintaining an efficient RTE system, as such a big database is problematic in terms of the memory usage and computational complexity. In this work, we show the processing time of a state-of-the-art logic-based RTE system can be significantly reduced by replacing its search-based axiom injection (abduction) mechanism by that based on Knowledge Base Completion (KBC). We integrate this mechanism in a Coq plugin that provides a proof automation tactic for natural language inference. Additionally, we show empirically that adding new knowledge data contributes to better RTE performance while not harming the processing speed in this framework.