Skip to content

Hypotheses

alexisantracoli edited this page Feb 3, 2022 · 5 revisions

Hypotheses

The sections below provide space for project members to discuss project expectations.

Hypothesis 1: NER and OCR will reveal useful information

We hypothesize that named-entity extraction from dirty OCR will yield information that enables archivists to discover people, places, and organizations that are not well represented in existing finding aids.

Hypothesis 2: Automated or semi-automated techniques will conform with MPLP

We hypothesize that we can develop automated or semi-automated processes that will require minimal effort from archivists.

Hypothesis 3: NER and OCR will make it possible to reveal connections between named entities that are obscured by current description

We hypothesize that by connecting named entities with finding aid components it will become possible to make connections between people, places, and organizations that are not currently described in the finding aids. (The expectation is that this will require additional manual metadata work.)