WEEK 4.1 Flashcards
(29 cards)
: What is Digital History?
Studying the past using digital tools like AI, OCR, and databases.
What was Cliometrics in the 1960s?
Using computers for statistical analysis of historical data.
: What does OCR do?
Turns printed text into machine-readable digital text.
What does HTR do?
Uses AI to read and transcribe handwritten documents.
What is a challenge with OCR/HTR on old texts?
Low-quality scans and archaic handwriting lead to mistakes.
What does Named Entity Recognition (NER) do?
Finds people, places, and dates in text.
Why is NER hard on historical texts?
Due to old spellings, obsolete names, and OCR errors.
What is semantic drift?
When word meanings change over time (e.g., “liberal” then vs. now).
What is network analysis in history?
Mapping relationships (e.g., between thinkers, politicians).
: What is Linked Open Data (LOD)?
A way to connect historical facts across multiple databases (e.g., Wikidata).
What is a Document Analysis System?
AI + NLP + computer vision tools that read and analyze documents.
What is an example of an HTR tool?
: Transkribus – reads handwriting in historical documents.
What does AI-based indexing do?
Extracts and organizes info like names, dates, and places from scanned docs
Why do historical documents confuse AI?
Due to inconsistent spelling, old styles, and faded ink.
What are Transformer-based OCR models?
New AI models (like BERT) that better understand text context and layout.
What is one ethical issue in digital history?
Who owns the data and how it’s used by AI models.
: How can historians improve AI results?
By correcting errors, training models, and checking for bias.
: What is semantic analysis in DAS?
Going beyond word recognition to understand meaning and context.
What does OCR stand for and do?
A) Open Code Recognition – Converts audio to text
B) Optical Character Recognition – Converts printed text into digital text
C) Optical Content Retrieval – Finds images in scanned documents
D) Old Character Rendering – Displays historic fonts
b
. Which AI tool is commonly used for recognizing handwritten historical texts?
A) DeepScribe
B) Transkribus
C) ChatScribe
D) TextFinder
b
What is one of the earliest forms of digital history from the 1960s?
A) Machine Translation
B) Cliometrics
C) Smart Archives
D) Digital Storytelling
b
. What does Named Entity Recognition (NER) extract from text?
A) Emotions and opinions
B) Fonts and formats
C) People, places, and dates
D) Charts and tables
c
What is a major challenge when applying NER to historical texts?
A) Lack of grammar rules
B) Use of poetic language
C) Spelling variations and obsolete names
D) Too much punctuation
c
What is the Semantic Web useful for in digital history?
A) Making web pages load faster
B) Connecting and linking historical data sets
C) Downloading digital books
D) Archiving images only
b
