Japanese PDF Annotation Expert for Precise Data Labeling

Obsidian; Zürich; Vor Ort; CHF 34.000 - 55.000; Vollzeit

Zusammenfassung

Mercor in Switzerland seeks meticulous Japanese language annotators to build training data for multilingual document understanding. You will label, transcribe, and structure real Japanese documents, including handwritten and multi-column pages.

You must be a native Japanese speaker with kanji, hiragana, katakana, furigana, and experience in precise transcription or editorial work. This role emphasizes character-level accuracy and systematic taxonomy across hundreds of pages.

Qualifikationen

• Native Japanese with full command of kanji, hiragana, katakana, furigana and variant forms.
• Experience in interpretation, transcription, translation, editorial work or similar document-intensive tasks.
• Character-level accuracy is essential in this role.

Aufgaben

• Source publicly accessible Japanese PDFs with multimodal elements and record provenance.
• Annotate structure: bound every meaningful region and assign type and reading order.
• Record relationships by linking regions to their parent figure or table.
• Transcribe text exactly as appears, including kanji, furigana and handwriting; flag illegible regions.
• Capture page metadata: language, document type, source, dimensions and flags for tables, formulas and handwriting.
• Review work via second-expert review to ensure accuracy and consistency.

Kenntnisse

Native Japanese speaker

Attention to detail

Systematic

Adaptable to unfamiliar layouts

Furigana and variant forms

Jobbeschreibung

Mercor in Switzerland seeks meticulous Japanese language annotators to build training data for multilingual document understanding. You will label, transcribe, and structure real Japanese documents, including handwritten and multi-column pages.

You must be a native Japanese speaker with kanji, hiragana, katakana, furigana, and experience in precise transcription or editorial work. This role emphasizes character-level accuracy and systematic taxonomy across hundreds of pages.