Skip to Main content Skip to Navigation
Conference papers

Entities as topic labels: Improving topic interpretability and evaluability combining Entity Linking and Labeled LDA

Abstract : In order to create a corpus exploration method providing topics that are easier to interpret than standard LDA topic models, here we propose combining two techniques called Entity linking and Labeled LDA. Our method identifies in an ontology a series of descriptive labels for each document in a corpus. Then it generates a specific topic for each label. Having a direct relation between topics and labels makes interpretation easier; using an ontology as background knowledge limits label ambiguity. As our topics are described with a limited number of clear-cut labels, they promote interpretability, and this may help quantitative evaluation. We illustrate the potential of the approach by applying it in order to define the most relevant topics addressed by each party in the European Parliament's fifth mandate (1999-2004).
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01483336
Contributor : Pablo Ruiz Fabo <>
Submitted on : Sunday, March 5, 2017 - 1:37:07 PM
Last modification on : Wednesday, February 19, 2020 - 8:30:02 PM

Links full text

Identifiers

Collections

Citation

Federico Nanni, Pablo Ruiz. Entities as topic labels: Improving topic interpretability and evaluability combining Entity Linking and Labeled LDA. Digital Humanities 2016, Jagiellonian University and Pedagogical University, Jul 2016, Cracovie, Poland. pp.632-635. ⟨hal-01483336⟩

Share

Metrics

Record views

133