Named-entity recognition
Also: NER
Tagging spans of text that refer to people, places, organisations, dates, and similar categories.
Last updated
NER identifies spans of text that refer to people, places, organisations, dates, and similar categories. A paragraph becomes a set of tagged mentions: “Dr. Huxley” → PERSON, “Somerset House” → LOCATION, “1891” → DATE.
Why it matters for your research. NER is the enabling step behind many DH questions — who appears with whom, how geographical focus shifts, how an archive describes groups of people. It is error-prone on historical sources (unfamiliar names, OCR noise, varied orthography) and those errors compound downstream, so reviewing NER output is itself part of the methodology.
In Archēglyph. On the 90-day roadmap (spaCy / Stanza). It will land as an analysis plugin and add entity facets to search plus a dataset-level entity summary.
Not to be confused with. Keyword search finds words; NER finds references — it disambiguates “Paris” the city from “Paris” the person in the same corpus.