Combining Axiom Injection and Knowledge Base Completion for Efficient Natural Language Inference
Combining Axiom Injection and Knowledge Base Completion for Efficient Natural Language Inference
复制标题
DOI：
10.1609/aaai.v33i01.33017410
复制
发表时间：
2018-11
期刊：
影响因子：
--
通讯作者：
Masashi Yoshikawa;K. Mineshima;Hiroshi Noji;D. Bekki
中科院分区：
文献类型：
--
作者：
Masashi Yoshikawa;K. Mineshima;Hiroshi Noji;D. Bekki
在基于逻辑的推理任务（例如识别文本蕴涵（RTE））中，系统拥有大量知识数据非常重要。然而，在添加更多知识数据以提高 RTE 性能和维护高效 RTE 系统之间需要权衡，因为这样的大数据库在内存使用和计算复杂性方面存在问题。在这项工作中，我们展示了通过将基于搜索的公理注入（溯因）机制替换为基于知识库补全（KBC）的机制，可以显着减少最先进的基于逻辑的 RTE 系统的处理时间。我们将此机制集成到 Coq 插件中，该插件为自然语言推理提供了证明自动化策略。此外，我们凭经验表明，添加新的知识数据有助于提高 RTE 性能，同时不会损害该框架中的处理速度。
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.