A5089571454- Topic Modeling
- Natural Language Processing Techniques
- Semantic Web and Ontologies
- Scientific Computing and Data Management
- Data Quality and Management
- Advanced Graph Neural Networks
- Research Data Management Practices
Universität Hamburg
2023
This paper presents a scholarly Knowledge Graph Question Answering (KGQA) that answers bibliographic natural language questions by leveraging large model (LLM) in few-shot manner. The initially identifies the top-n similar training related to given test question via BERT-based sentence encoder and retrieves their corresponding SPARQL. Using question-SPARQL pairs as an example creates prompt. Then pass prompt LLM generate Finally, runs SPARQL against underlying KG - ORKG (Open Research KG)...
This paper introduces a scholarly Question Answering (QA) system on top of the NFDI4DataScience Gateway, employing Retrieval Augmented Generation-based (RAG) approach. The NFDI4DS as foundational framework, offers unified and intuitive interface for querying various scientific databases using federated search. RAG-based QA, powered by Large Language Model (LLM), facilitates dynamic interaction with search results, enhancing filtering capabilities fostering conversational engagement Gateway...
Existing Scholarly Question Answering (QA) methods typically target homogeneous data sources, relying solely on either text or Knowledge Graphs (KGs). However, scholarly information often spans heterogeneous necessitating the development of QA systems that can integrate from multiple sources. To address this challenge, we introduce Hybrid-SQuAD (Hybrid Dataset), a novel large-scale dataset designed to facilitate answering questions incorporating both and KG facts. The consists 10.5K...
©NFDI4DataScience (NFDI4DS) is a consortium to support researchers in all stages of the research data lifecycle conduct their line with FAIR principles. The developed infrastructure targets from wide range disciplines science and AI. We present ideas NFDI4DS gateway portal. Two approaches navigate digital objects (articles, data, machine learning models, workflows, scripts/code, etc.) various resources such as ORKG, DBLP database, other knowledge graphs (KGs). Transparency, reproducibility,...