World Publishing Houses — Publishing Intelligence Dataset

Problem

International publishing metadata is fragmented across retailers, publishers, libraries, rights catalogs, award announcements, and market news. A reader may see that a book exists, but not whether it is available in English, who translated it, which publisher handled the edition, or whether the attribution is reliable.

For translators, publishers, bookstores, and rights professionals, the same fragmentation makes it difficult to compare national markets, identify translation gaps, track rights movement, and understand where acquisition opportunities may exist.

Trust and verification

The dataset separates evidence levels instead of flattening every row into the same confidence level. This matters because publishing metadata can be incomplete, inconsistent, or source-dependent.

Verified public-source rows can support reader-facing confidence. Curated rows are useful for portfolio demonstration, research planning, and rights-watchlist exploration, but they need manual review before being treated as authoritative.

Trust fields

• verification_status
• source_url
• source_name
• translation_path
• translator
• original_language
• translated_language
• rights_signal
• reader_bucket

Reader mode

• Find books available in English now.
• Discover upcoming translations.
• Explore books not yet available in English.
• Browse by country, publisher, or translator.

Professional mode

• Track rights and acquisition signals.
• Monitor publisher activity.
• Identify translation gaps.
• Compare countries and market coverage.
• Follow market events and awards.

Product insight

This dataset can support country pages, publisher pages, translated book discovery, translator attribution pages, rights dashboards, bookstore buying-intelligence views, trust badges, and translation-chain transparency.

For machine learning work, it also creates natural opportunities for entity resolution, classification, metadata conflict detection, embeddings-based search, and recommendation workflows.

My role

I designed this as a product-minded data portfolio asset: product researcher, data modeler, QA reviewer, dashboard designer, and ML/Data Science portfolio builder.

I planned the data structure, researched source-backed examples, separated verified and curated records, and mapped how the dataset could support both reader-facing and professional workflows.