Build machine-learning training datasets from historical genealogical documents.
Decipher historical documents and annotate language data for NLP datasets.
Create text-area relationship hierarchies and map document layouts.
Curate structured and unstructured data, review datasets for errors, and make corrections.
Summarised from the posting by AI and can be wrong. The full description is the source.
Requirements
Experience
2 years of relevant or related experience, or equivalent experience
Experience with Hiragana, Katakana, Kanji, and Furigana
Experience working with historical documents
Education
BA/BS in Computer Science, Linguistics, Family History, or another bachelor's degree with related or equivalent experience
Skills
Native-level fluency in Japanese
Expertise in Japanese cursive scripts
Business-level fluency in English
Demonstrated paleography skills to accurately decipher historical documents
Demonstrated linguistic skills to build NLP datasets
High proficiency in Microsoft Word, PowerPoint, and Excel
Experience with additional languages (preferred)
Full job description
From Church’s careers site
This is a 2 year Fixed-Term position. Seeking an individual with a deep knowledge of and demonstrated native-level Japanese language expertise reading historical genealogical documents in Hiragana, Katakana, Kanji, Furigana, and cursive scripts (Click HERE to see sample documents) to help us build high-quality training data for machine learning systems. This person’s work will help make billions of records available in FamilySearch’s automation and machine learning platform.
Exercise demonstrated expertise in paleography, linguistics, technology, and historical records to build machine learning training datasets for historical genealogical documents
Show full descriptionShow less
Use paleography skills to accurately decipher historical documents
Annotate language data with linguistic information to build natural language processing (NLP) datasets
Model how individuals naturally read historical documents by creating hierarchies of relationships between areas of text
Precisely map the layout of historical documents
Perform data curation and quality assurance by:
Curating large volumes of structured and unstructured data
Reviewing datasets for errors and providing timely corrections
Completing other data modeling activities and duties as assigned
Enable FamilySearch’s automation efforts by meeting aggressive deadlines and delivering consistently high output, quality, and accuracy
Education:
BA/BS in Computer Science, Linguistics, Family History, or other bachelor’s degree with related or equivalent experience required
Experience:
2 years relevant or related experience or equivalent experience
Skills & Abilities:
Native level fluency in Japanese including:
Expertise in Japanese cursive scripts (To see samples, click HERE)
Experience with Hiragana, Katakana, Kanji, and Furigana
Experience with additional languages a plus
Business level fluency in English
Demonstrated paleography skills to accurately decipher historical documents
Demonstrated linguistic skills to build NLP datasets
Experience working with historical documents
Strong technical and analytical aptitude with a passion for data, efficiency, and accuracy
Independent worker who is self-motivated, dependable, detail oriented, responsible, self-disciplined, and a team player with a record of timely delivery of requests
Willingness to support several projects at one time, and to accept reprioritization as necessary in a fast paced, constantly evolving environment
Comfortable handling a high volume of work on a daily basis
High proficiency in Microsoft Office tools including: Word, PowerPoint, and Excel