Researchers have developed a two-phase entity resolution pipeline to link historical narrator identities across disparate Arabic biographical databases. The system first matches narrator names from the Sanadset 650K corpus to the Hawramani Arabic Lexicon, achieving over 51% linkage. Subsequently, it cross-references Hawramani data with the Muslimscholars database using a multi-signal function that combines name similarity, death year proximity, and reliability grades, resulting in a 94.7% linkage rate. This process enables the creation of a comprehensive transmission graph of Islamic Hadith literature, enriched with cross-source biographical metadata, and the release of annotated link corpora and the graph as open resources. AI
IMPACT This research advances entity resolution techniques, potentially applicable to other historical or fragmented datasets.
RANK_REASON The cluster describes a research paper detailing a novel entity resolution pipeline for historical databases.
- Hadith
- Hadithtransmitters
- Hawramani Arabic Lexicon
- Muslimscholars
- Sanadset 650K
- alphaXiv
- arXiv
- CatalyzeX
- DagsHub
- Gotit.pub
- Hugging Face
- Islamic Hadith
- ScienceCast
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
