利用路徑差之路徑排序以保持知識庫完整性

Path Ranking with Path Difference Sets for Maintaining Knowledge Base Integrity

黃博政(Po-Cheng Huang)

指導教授 ： 陳信希

國立臺灣大學/電機資訊學院/資訊工程學研究所/碩士(2018年)

https://doi.org/10.6342/NTU201803782

摘要

以往有不少研究在於讓知識庫自動補全其缺漏的知識，而知識會隨著時間而改變，在知識庫的新增或刪除操作可能會導致其它的相關事實變成錯誤，否則將導致知識庫內部的矛盾。因此我們必須自動的將知識庫內部發生衝突的事實和過期事實移除，如此才能保持知識庫的完整性。我們提出透過整合曾經被刪除的事實，讓我們的模型可以偵測這些過期與衝突的事實以保持知識庫沒有雜訊。一個乾淨的知識庫將可以有效地提供後續的應用像是問答系統或是知識庫補全。

關鍵字

知識庫補完 ； 知識庫完整性 ； 路徑排序

並列摘要

Knowledge base completion involves in discovering missing facts. However, knowledge changes over time. The issue of knowledge base integrity arises when the operations for knowledge base update including insertion and deletion result some facts inconsistent with others in the knowledge base. Therefore, the removal of out-of-date facts is required to keep knowledge base integrity. In this study, we explore two kinds of approaches to knowledge base integrity. In this way, we can eliminate the conflicting facts and out-of-date facts. Our work provides a significant benefit for other tasks such as question answering and knowledge base completion.

並列關鍵字

knowledge base completion ； knowledge base integrity ； path ranking

參考文獻

1. Auer, S., et al., Dbpedia: A nucleus for a web of open data, in The semantic web. 2007, Springer. p. 722-735.

2. Bollacker, K., et al. Freebase: a collaboratively created graph database for structuring human knowledge. in Proceedings of the 2008 ACM SIGMOD international conference on Management of data. 2008. AcM.

3. Bordes, A., et al. Translating embeddings for modeling multi-relational data. in Advances in neural information processing systems. 2013.

4. Frank, J.R., et al., Building an entity-centric stream filtering test collection for TREC 2012. 2012, MASSACHUSETTS INST OF TECH CAMBRIDGE.

5. Gardner, M. and T. Mitchell. Efficient and expressive knowledge base completion using subgraph feature extraction. in Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing. 2015.