ediewilson/ttu-encoder

DDHI Encoder and Training Tools development

A collection of command-line utilities to assist in the creation of TEI-encoded oral history interviews for the Dartmouth Digital History Initiative.

Installation

Use pip to install this package:

pip install ttu-encoder

To peform named-entity tagging with ttu_tag, you will need a Spacy model. Before running ttu_tag, install Spacy's small English model:

python -m spacy download en_core_web_sm

See the Spacy documentation for more information.

Use

Use ttu_convert to transform a DOCX-encoded transcription into a simply structured TEI document.

ttu_convert ~/Desktop/transcripts/zien_jimmy_transcript_final.docx -o tmp.tei.xml

Use ttu_tag to add named-entity tags to a TEI-encoded transcription:

ttu_tag -o zien.tei.xml tmp.tei.xml

Encoders are then expected to edit the text of the interview, correcting automatically generated named-entity tags and adding new ones.

Use ttu_generate_standoff to create a <standOff> element in the interview and link the entities to names in the text.

Use ttu_mentioned_places to extract the places in a TEI file's standoff markup and print it as tab-separated values:

ttu_mentioned_places lovely.tei.xml > lovely.tsv

Then use OpenRefine or another tool to refine this list with identifiers and other metadata.

Use ttu_update_places to update the places in a TEI file's standoff markup with identifiers and geo-coordinates obtained via OpenRefine or other procedure:

ttu_update_places lovely.tei.xml lovely_updates.tsv >
updated_lovely.tei.xml

Similarly, use ttu_mentioned_events and ttu_update_events to perform the same operations for events.