@univ-lehavre/atlas-openalex

Interactive CLI for OpenAlex data mining. Guides a researcher through author search, affiliation selection, and publication curation — backed by DuckDB, ML embeddings, and string similarity grouping.

Usage

pnpm -F @univ-lehavre/atlas-openalex start
# or with a name argument
node --env-file .env dist/index.js --name "Dupont Jean"

The CLI:

1. Searches OpenAlex for authors matching the given name
2. Selects relevant display name variants
3. Filters by institution affiliation
4. Curates the final list of publications

Configuration

Requires a .env file:

OPENALEX_API_URL=https://api.openalex.org
PER_PAGE=25
OPENALEX_API_KEY=          # optional

Internals

src/fetch/

Export	Description
searchAuthors(name)	Searches OpenAlex authors by display name
retrieve_articles(authorIds, institutionIds)	Fetches articles for given authors and institutions
retrieve_articles_given_work_ids(workIds)	Fetches articles by OpenAlex work IDs

src/group/

String similarity utilities for grouping affiliation strings:

Export	Description
groupBySimilarity(strings, threshold)	Groups strings by Levenshtein similarity
groupBySimilarityWithScore(strings, threshold)	Same, with per-group similarity score
groupByNGramSimilarity(strings, threshold)	Groups strings by n-gram (bigram/trigram) Jaccard similarity
normalizeString(input, options)	Normalizes a string (diacritics, punctuation, case)