[1O4-GS-4] Knowledge utilization and sharing: Semantic web (1)

[1O4-GS-4-01] Report on the Second Knowledge Graph Reasoning Challenge 2019

- Toward the eXplainable AI System -

〇Takahiro KAWAMURA^1,8, Shusaku Egami², Kyoumoto Matsushita³, Koutarou Tamura^4,5, Mitsuhiro Sumita⁴, Yasunori Hokazono⁴, Mori Kurokawa⁶, Takanori Ugai³, Kouji Kozaki⁷ (1. National Agriculture and Food Research Organization, 2. Electronic Navigation Research Institute, 3. Fujitsu Laboratories Ltd., 4. Nomura Research Institute, Ltd., 5. NRI digital, Ltd., 6. KDDI Research, Inc., 7. Osaka Electro-Communication University, 8. Japan Science and Technology Agency)

Keywords:Knowledge Graph, Explainability, Machine Learning, Reasoning

Results of the second knowledge graph reasoning challenge 2019 are reported in this paper.
According to applying AI systems to a wide variety of social systems, the explanation for the AI decision is becoming important to guarantee the security and safety of such AI systems.
Therefore, the Special Interest Group on Semantic Web and Ontology of JSAI started this challenge in 2018, which calls for techniques for reasoning and/or estimating criminals with a reasonable explanation based on knowledge graphs representing well-known stories of Sherlock Holmes.
This paper describes a summary report of the second challenge, an approach based on Abduction and Answer Set Programming, and an approach using Graph Embedding, then their results are evaluated.
Finally, we conclude this paper with the plans and issues for the next challenge in 2020.