L-KD is a tool that relies on available linguistic and knowledge resources to perform keyphrase clustering and labelling. The aim of L-KD is to help finding and tracing themes in English and Italian text data, represented by groups of keyphrases and associated domains.

The tool takes advantage of the availability of Keyphrase Digger (KD), a multilingual rule-based system that detects a weighted list  of  n-grams  representing  the  most important concepts in a text. These key-concepts are then linked to WordNet Domains in order to create clusters of key-concepts labelled by domain. The problem of ambiguous concepts, i.e. possibly belonging to more than one WordNet domain, is tackled by using ConceptNet 5, a multilingual knowledge source containing single and multi-word concepts linked to each other by a broad set of relations covering different types of associations.  


Moretti, G., Sprugnoli, R., Tonelli, S. KD Strikes Back: from Keyphrases to Labelled Domains Using External Knowledge Sources. In Proceedings of the Third Italian Conference on Computational Linguistics (CLiC-it 2016), Naples, Italy.