Data Sources

Overview

The project uses Nobel Prize and academic publication data from multiple sources.

Primary Dataset (data/26963326/)

The core dataset contains structured data about Nobel Prize laureates and their scientific publications, sourced from OpenAlex and Nobel Prize archives.

CSV Tables (data/26963326/db_data/)

laureate.csv — Nobel Laureates

Field	Type	Description
laureate_id	int	Unique laureate identifier
name	string	Full name
gender	string	Gender
nationality	string	Country of nationality
education	string	Educational background
wikipedia_url	string	Wikipedia page URL
wikidata_id	string	Wikidata entity ID

Records: 757 laureates

award_info.csv — Award Details

Field	Type	Description
laureate_id	int	Laureate identifier
year	int	Award year
category_id	int	Category identifier
motivation	string	Award motivation text
prize_amount	float	Prize amount

Records: 761 awards

nobel_prize_category.csv — Prize Categories

Field	Type	Description
category_id	int	Category identifier
category_name	string	Category name

Categories: Physics, Chemistry, Physiology/Medicine, Economics

work.csv — Publications

Field	Type	Description
openalex_work_id	string	OpenAlex work identifier
title	string	Paper title
keywords	string	Semicolon-separated keywords
abstract_inverted_index	JSON	OpenAlex inverted index format
referenced_works	string	Cited work IDs
doi	string	Digital Object Identifier
publication_year	int	Publication year

Records: ~245,000 papers

work_authorship.csv — Authorship

Field	Type	Description
openalex_work_id	string	Work identifier
author_id	string	Author identifier
institution_id	string	Institution identifier
position	string	Author position (first, middle, last)

Records: ~1.67 million relationships

work_citation_by_year.csv — Citation Counts

Field	Type	Description
openalex_work_id	string	Work identifier
year	int	Citation year
cited_by_count	int	Number of citations in that year

Records: ~9 million entries

Other Tables

• author.csv — ~297K OpenAlex author IDs
• institution.csv — ~13K institutions (name, ROR, geo-location)
• source.csv — ~11.5K journals/publishers
• laureate_openalex_matching.csv — ~840 laureate ↔ OpenAlex ID matches
• laureate_location_info.csv — ~2K laureate geo-information

JSON Data (data/26963326/json/)

publication_records.json (2.3 GB)

Line-delimited JSON (JSONL) containing full metadata for 253K papers:

{
  "id": "W2078536640",
  "title": "The ubiquitin-proteasome proteolytic pathway",
  "publication_year": 1998,
  "cited_by_count": 1250,
  "authorships": [...],
  "referenced_works": ["W...", "W..."],
  "abstract_inverted_index": {...},
  "keywords": [...]
}

laureate_info.json (924 KB)

Laureate details with geographic coordinates for birth, award, and death locations.

award_details.json (664 KB)

Award details including geographic coordinates.

External API Sources

OpenAlex API

• Purpose: Supplement paper concepts, topics, and field classifications
• Endpoint: https://api.openalex.org
• Cache: output/openalex_cache/
• Documentation: OpenAlex API Docs

Semantic Scholar API

• Purpose: Fetch abstracts and TLDR summaries
• Cache: output/openalex_cache/semantic_scholar/
• Documentation: S2 API Docs

Unpaywall API

• Purpose: Find Open Access PDF URLs
• Cache: output/openalex_cache/unpaywall/
• Documentation: Unpaywall API Docs

Data Coverage

Signal	Coverage	Notes
Paper titles	~100%	Available for nearly all papers
Keywords	~70%	Semicolon-separated keyword strings
Abstracts (inverted index)	~57%	OpenAlex format, can be decoded
Citation network	Most	Referenced works available
Year-by-year citations	~9M records	Comprehensive citation tracking

Known Gaps

Missing Data	Impact	Mitigation
Paper full text	Cannot extract deep concept details	Use abstracts + LLM augmentation
Subject classification	Cannot directly categorize papers	OpenAlex API + LLM inference
Concept ontology	No pre-built hierarchy	Build via LLM + seed concepts
Cross-field annotations	No direct labels	Citation network analysis + LLM