Similarity Retrieval Algorithm for Virtual Knowledge Graph
Project/Area Number |
22K18004
|
Research Category |
Grant-in-Aid for Early-Career Scientists
|
Allocation Type | Multi-year Fund |
Review Section |
Basic Section 62020:Web informatics and service informatics-related
|
Research Institution | Japan Advanced Institute of Science and Technology |
Principal Investigator |
RACHARAK Teeradaj 北陸先端科学技術大学院大学, 先端科学技術研究科, 講師 (30847512)
|
Project Period (FY) |
2022-04-01 – 2025-03-31
|
Project Status |
Granted (Fiscal Year 2022)
|
Budget Amount *help |
¥3,900,000 (Direct Cost: ¥3,000,000、Indirect Cost: ¥900,000)
Fiscal Year 2024: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2023: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
Fiscal Year 2022: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
|
Keywords | Virtual Knowledge Graph / Similarity Algorithms / Information Retrieval / Explainable Similarity / Description Logic / Analogical Reasoning / Knowledge Graph |
Outline of Research at the Start |
Virtual knowledge graph (VKG) is an emerging research area that has received huge attention in recent years for integration and access of various databases. However, existing VKGs do not support the users to query similar information due to lack of similarity algorithms. Thus, we aim to develop novel algorithms based on Description Logics that can advance the querying of VKGs with explainable similarity, called "VKG-SIM". We envision that our proposed VKG-SIM will lay a solid ground for developing AI-based integration technologies that enable the utilization of diverse and massive data.
|
Outline of Annual Research Achievements |
We have investigated algorithms based on Description Logics to advance the querying of VKG with an explainable similarity score and obtained outlines of the algorithms. We also studied algorithms that go beyond Description Logics. For example, how to inference with analogical reasoning between terminologies, how to complete knowledge graph with weighted knowledge graph embedding. These results can be used to support the implementation in our next step.
|
Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
We have obtained similarity algorithms for Description Logics. We are also conducting a preliminary study of using the algorithms for querying in our VKG system. Regarding our publication, we have published a conference paper at ICTAI, which is a leading IEEE-CS annual scientific meeting for three decades. In addition, we have submitted a journal paper to IEEE Access (Q1; under review) and another journal paper to SWJ (Q1; under review).
|
Strategy for Future Research Activity |
As for our next step, we will mathematically prove our algorithm in order to conclude our mathematical framework for the VKG. At the same time, we are under the implementation of the investigating framework. The goal of this year is to ensure both theoretical and practical correctness of the system.
|
Report
(1 results)
Research Products
(1 results)