Competencies mining with LookinLabs

William Dedzoe (1) , Jean Hany (1, 2) , Albert Benveniste (1, 2)

1 Inria Rennes – Bretagne Atlantique
2 HYCOMES - Modélisation hybride & conception par contrats pour les systèmes embarqués multi-physiques

LookinLabs is a tool that returns teams, individuals, and publications best semantically correlated with a query, submitted as a list of keywords or a text. In short, it performs competencies mining. The teams, individuals, and publications must belong to a defined institution or lab (e.g., Inria). The tool relies on the HAL open archive, complemented by administrative information bases offered by the institution or lab. This side information turns out to be essential for the quality of the responses, as meta-data (author's names and affiliations) are subject to many variations. To the best of our knowledge, no similar tool or service exists in France.

Mots clés: competencies mining; text mining; Elasticsearch; HAL

Domaines: Traitement du texte et du document; Recherche d'information [cs.IR]

William Dedzoe, Jean Hany, Albert Benveniste. Competencies mining with LookinLabs. [Research Report] 9158, Inria Rennes Bretagne Atlantique. 2018, pp.1-27. ⟨hal-01739845v1⟩