topic_refinement: Computes the refinement score of a topic

In lasy/alto: Aligns Topics across LDA Models

Computes the refinement score of a topic

Description

The refinement score of a topic reflects the average weight that each of this topic's children attributes to that topic. A topic with a high refinement score is a topic for which all of the subsequent topics connected by some forward weight to this topic have high backward weights to that topic.

Usage

topic_refinement(model, topic, aligned_topics)

Arguments

model	The index of the model containing the topic whose coherence we want
topic	The index of the topic whose coherence we want, within the current model
aligned_topics	An alignment object on which the coherence scores should be computed.

lasy/alto documentation built on June 23, 2024, 6:45 a.m.