KeyCloud is a simple keyphrase extractor and a keyphrase-cloud generator.
Is suitable to work with text corpora. It identifies each document keyphrases and generates a keyphrase-cloud that informs about the most scored keyphrases within the corpus.
KeyCloud contains several models:
tf-idf: A baseline model that uses tf-idf information to score the keyphrase candidates.
H1: A model based in simple heuristic, based in each document features only, to score the keyphrase candidates. (working paper)
H2: The same as H1 with idf information added.
EmbedRank s2v. A model based in the EmbedRank model.
If you use it please cite the following paper:
Won, Miguel, Bruno Martins, and Filipa Raimundo. Automatic extraction of relevant keyphrases for the study of issue competition. Proceedings of the 20th international conference on computational linguistics and intelligent text processing, Berkeley, La Rochelle, France. 2019.