A5006724001- Semantic Web and Ontologies
- Natural Language Processing Techniques
- Mathematics, Computing, and Information Processing
- Fault Detection and Control Systems
- Scientific Computing and Data Management
- Advanced Text Analysis Techniques
- Advanced Control Systems Optimization
- Advanced Data Processing Techniques
- Topic Modeling
- Distributed and Parallel Computing Systems
- Rheology and Fluid Dynamics Studies
- Advanced Statistical Process Monitoring
- Mineral Processing and Grinding
- Machine Learning in Materials Science
- Software Engineering Research
- High-Velocity Impact and Material Behavior
- Process Optimization and Integration
- Metallurgy and Material Forming
- Domain Adaptation and Few-Shot Learning
- Semiconductor materials and devices
- Electron and X-Ray Spectroscopy Techniques
- Solidification and crystal growth phenomena
- Iron and Steelmaking Processes
- Aluminum Alloys Composites Properties
- Advanced Database Systems and Queries
Kyoto University
2020-2025
Kyoto College of Graduate Studies for Informatics
2023
Chubu University
2021
Ritsumeikan University
2014
Digital twins are expected to play a pivotal role in digital transformation. Although process informatics has attracted much attention, physical models essential realizing the twins. However, building model of an industrial takes toil. We aim facilitate by developing automated AI, named AutoPMoB, which performs five tasks: 1) retrieving documents about target from literature databases, 2) converting format each document HTML format, 3) extracting information required for documents, such as...
Rheology plays a pivotal role in understanding and predicting material behavior by discovering governing equations that relate deformation stress, known as constitutive equations. Despite the critical importance of dynamics complex fluids, systematic methodology for deriving these from available data has remained significant challenge field. To overcome problem, we propose novel method named Rheo-SINDy, which employs sparse identification nonlinear (SINDy) models rheological data. Rheo-SINDy...
Rheology plays a pivotal role in understanding the flow behavior of fluids by discovering governing equations that relate deformation and stress, known as constitutive equations. Despite importance these equations, current methods for deriving them lack systematic methodology, often relying on sense physics incurring substantial costs. To overcome this problem, we propose novel method named Rheo-SINDy, which employs sparse identification nonlinear dynamics (SINDy) algorithm models from...
The extraction of variable definitions from scientific and technical papers is essential for understanding these documents. However, the characteristics definitions, such as length words that make up definition, differ among fields, which leads to differences in performance existing methods across fields. Although preparing training data specific each field can improve methods, it costly create high-quality data. To address this challenge, study proposes a new method generates definition...
Mathematical formulas are essential tools for conveying mathematical concepts. Definitions of symbols in often vary among different documents; thus, knowing the definitions is fundamental to grasping semantics formulas. This research targets how extract representing variables from documents on chemical processes. We defined three features focusing unique usage variable and these proposed a new definition extraction method. compared performance method with that representative conventional...
In industrial scenarios, the difficulty of collecting a sufficient amount data for analysis is long-standing problem. Transfer learning (TL) believed to be potential solution this TL methodology that transfers useful knowledge from previous and/or tasks in source domain help build more accurate model target domain. So far most research conducted fields computer vision and natural language processing, thus images or text are taken as inputs. study, we aim propose method scenarios;...
Heart rate variability (HRV) analysis plays an essential role in healthcare. HRV features cannot be extracted accurately from the R-R interval (RRI) when RRI data contains artifacts. Previous research for modifying with artifacts considered premature atrial contraction (PAC) and ventricular (PVC), which are most common types of extrasystoles occurring every day healthy persons. This proposed three new modification algorithms PAC PVC using nearest neighbor search (NNS) algorithms: k-nearest...
Physical model building is essential for realizing digital twins in the manufacturing industry and requires much toil. We aim to develop automated physical builder (AutoPMoB) that can automatically build models from literature databases. AutoPMoB several fundamental technologies, domain-specific datasets play a vital role developing such technologies. Although related variables have been created, there has no dataset chemical engineering domain. To create dataset, this study, we developed an...
Mathematical formulas play a prominent role in science, technology, engineering, and mathematics (STEM) documents; understanding STEM documents usually requires knowing the difference between equation groups containing multiple equations. When two can be transformed into same form, we call equivalent. Existing tools cannot judge equivalence of groups; thus, develop an algorithm to such using computer algebra system. The proposed first eliminates variables appearing only either group. It then...
The pharmaceutical industry is facing a transition from batch manufacturing to continuous in order overcome the inefficiencies of manufacturing. However, quality assurance in-process materials during remains an arduous task. Product pharmaceuticals assured by sustaining critical material attributes (CMAs) within region known as design space (DS). This study proposes method determine setpoints CMAs DS for line more than two processes, each with multiple CMAs. proposed based on distribution...
The Czochralski (CZ) process, a well-established monocrystalline silicon ingot production is nonlinear, time-varying batch process. In the semiconductor industry, it desirable to improve control method and manufacture higher-quality 300-mm-diameter ingots at lower cost. authors developed nonlinear model predictive based on gray-box (GB) of CZ process successive linearization. This study proposes for updating prediction model, handle plant-model mismatch. proposed constructs several GB models...
Abstract Mathematical formulas play a prominent role in science, technology, engineering, and mathematics (STEM) documents; understanding STEM documents usually requires knowing whether equation groups containing multiple equations are equivalent. However, existing tools cannot judge the equivalence of two groups; thus, we develop an algorithm to such using computer algebra system. The proposed first eliminates variables appearing only either group. It then checks one by one. If each group...