Modeling word burstiness using the Dirichlet distribution

By Rasmus Elsborg Madsen, David Kauchak, Charles Peter Elkan. International Conference on Machine Learning, 2005

330 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/dd5768b0605339c5e3b670ffbc2423018ed72c81

How researchers see this paper

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

In Artificial Intelligence

Citations in the corpus

Optimal Thresholding of Classifiers to Maximize F1 Measure (2014, methodology): “We apply a tf-idf text preprocessing step to the bag of words representation to account for word burstiness [10] and to elevate the impact of rare words.”
Author Manuscript, Published in "ieee Conference on Computer Vision and Pattern Recognition on the Burstiness of Visual Elements (unknown): “Burstiness has recently been shown to improve performance in the context of text classification [10] and text clustering [3].”
PyMC: a modern, and comprehensive probabilistic programming framework in Python (2023, background): “Models like this one are important in a variety of areas, including natural language processing ( Madsen, Kauchak & Elkan, 2005 ), ecology ( Harrison et al., 2019 ), and Genomics ( Holmes, Harris & Quince, 2012 ; Nowicka & Robinson, 2016 ).”
A General Method for Robust Bayesian Modeling (2015, methodology): “This is called a Dirichlet-multinomial compound distribution (Madsen et al., 2005).”
Introduction to the Dirichlet Distribution and Related Processes (2010, background): “One application area where the Dirichlet has proved to be particularly useful is in modeling the distribution of words in text documents [9].”
Evaluating the Stability of Embedding-based Word Similarities (2018, background): “Documents are often characterized by “bursty” words that are locally frequent but globally rare (Madsen et al., 2005), such as the name of a defendant in a court case.”
A dirichlet multinomial mixture model-based approach for short text clustering (2014, background): “However, this is a good property in the text clustering problem, because words in a document tend to appear in bursts: if a word appears once, it is more likely to appear again [14],[23].”
Topics over time: a non-Markov continuous-time model of topical trends (2006, unknown): “One could estimate the values of the hyper-parameters of the TOT model, α and β , from data using a Gibbs EM algorithm [1].”
SIFT Meets CNN: A Decade Survey of Instance Retrieval (2016, background): “ Although extensive studies on TF-IDF have been conducted in text analysis, relatively fewer are witnessed in image retrieval. An important problem associated with visual word weighting is burstiness [130], [131], [132]. This refers to the phenomenon whereby repetitive structures appear in an image:”