Automatic expansion of lexicon for organic residues valorization in emerging and developing countries

A previously created lexicon (https://doi.org/10.18167/DVN1/HNZZSI) was processed to: combine existing synonymic terms into consolidated relevant terms, and extract additional morphosyntactic variants from a corpus of 7,692 publications (titles and abstracts). The extraction of variants was based on the Fastr method (Jacquemin C., 1994), a driven linguistic approach that enables the extraction of term variants from full-text documents. The dataset contains two types of files: "XX_consolidated": these files contain the lexicons built for three lists of relevant terms (OWT, TM and AV) after automatic consolidation with Fastr (sheet raw) and after manual validation (sheet validated). Terms are split into two columns: seed_terms and variants. The variants column contains terms from the lexicon that were considered synonyms of the corresponding seed_term. "XX_variants_from_corpus": these files contain the lexicons' seed terms (from their consolidated version) along with the morphosyntactic variants extracted using the Fastr method. The code is openly accessible at: https://github.com/SarahVal/FastrOrganicWastes/tree/main. __________________________________________________________________ Jacquemin C. Fastr: a unification-based front-end to automatic indexing. In: Intelligent multimedia information retrieval systems and management - Volume 1, Le centre de hautes études internationales d’informatique documentaire, Paris, FRA, RIAO ’94, pp. 34–47, 1994