Evaluating Topic Coherence Using Distributional Semantics

By Nikolaos Aletras, Mark Stevenson. International Conference on Computational Semantics, 2013

355 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/bcd1c98f58aa580a628cf4c690efca947f89dbae

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

An analysis of the coherence of descriptors in topic modeling (2015, methodology): “TC-W2V – as an analog to the DS measures of Aletras and Stevenson (2013), we propose the creation of term vectors w v using a word2vec model (Mikolov et al., 2013).”
Topic Modeling Using Latent Dirichlet allocation (2021, background): “The aforementioned research effort advocates that the quality of a topic is often determined by the coherence [1, 101, 145] of its constituent words and its relative importance [1, 69] to the analysis task in comparison to other topics [70].”
Topically Driven Neural Language Model (2017, background): “Newman et al. (2010b); Mimno et al. (2011); Aletras and Stevenson (2013) propose automatic approaches to computing topic coherence, and Lau et al. (2014) summarises these methods to understand their differences.”
Measuring Model Biases in the Absence of Ground Truth (2021, methodology): “This information theoretic approach to uncovering word associations became a prominent method in the field of Natural Language Processing, with applications ranging from measuring topic coherence [1] to collocation extraction [4] to great effect, although often requiring a good deal of preprocessing”
Full-Text or Abstract? Examining Topic Coherence Scores Using Latent Dirichlet Allocation (2017, background): “The differences between topics are calculated with a topic coherence measure [18]–[21] that shows, in contrast to the likelihood of heldout data, a higher correlation with human topic ranking data, the gold standard for topic interpretability.”
Machine Reading Tea Leaves: Automatically Evaluating Topic Coherence and Topic Model Quality (2014, methodology): “Aletras and Stevenson (2013a) introduced distributional semantic similarity methods for computing coherence, calculating the distributional similarity between semantic vectors for the top-N topic words using a range of distributional similarity measures such as cosine similarity and the Dice…”
Exploring the Space of Topic Coherence Measures (2015, methodology): “In case W ′ and W ∗ consist of single words, the vector elements are just the direct confirmations as suggested in [1].”