Lexicon for Organic Residue Valorization from 7,692 Publications

Delivering a lexicon for organic residue valorization, automatically expanded from a corpus of 7,692 scientific publications using the Fastr linguistic method. It contains two file types: consolidated lexicons with seed terms and synonym variants, and files listing morphosyntactic variants extracted from the corpus. The work was authored by Sarah Valentin and the code is openly accessible.

Use Cases

• Apply the Fastr method to extract additional morphosyntactic variants for seed_terms from new text corpora.
• Analyze synonymic relationships in the lexicon by linking seed_terms with their corresponding variants for term consolidation.
• Validate and refine the manually validated lexicons in the 'validated' sheets for specific term lists like OWT, TM, and AV.
• Build a text mining pipeline for organic waste literature using the provided seed_terms and variants as a knowledge base.

Strengths

• The lexicon is derived from a substantial corpus of 7,692 scientific publications.
• The dataset includes both automatically consolidated and manually validated versions of lexicons for three distinct term lists (OWT, TM, AV).
• The extraction method, Fastr, is a defined linguistic approach for extracting term variants from full-text documents.

Limitations

• The dataset scope is limited to titles and abstracts from publications, not full texts.
• The underlying corpus and geographic focus of the publications are not specified, which may introduce thematic or regional bias.
• The size, row count, and specific file structures of the lexicons are unknown, making it difficult to assess scale and complexity.

Provenance

Source

CIRAD Harvested Collection, authored by Sarah Valentin.

Collection Method

A previously created lexicon was processed using the Fastr linguistic method to consolidate synonyms and extract morphosyntactic variants from a corpus of 7,692 publications.

Freshness

The dataset record was last updated on 2026-01-24.