Publication result detail

Latent Dirichlet Allocation (LDA) Approximation Analysis of Financial-Related Text Messages

JANKOVÁ, Z.

Original Title

Latent Dirichlet Allocation (LDA) Approximation Analysis of Financial-Related Text Messages

English Title

Latent Dirichlet Allocation (LDA) Approximation Analysis of Financial-Related Text Messages

Type

WoS Article

Original Abstract

Topic modeling is one of the most widely used NLP techniques for discovering hidden patterns and latent relationships in text documents. Latent Dirichlet Allocation (LDA) is one of the most popular methods in this field. There are different approaches to tuning the parameters of LDA models such as Gibbs sampling, variational method, or expected propagation. This paper aims to compare individual LDA parameter tuning approaches with respect to their performance and accuracy on textual data from the financial domain. From our results, it can be concluded that for text datasets obtained from financial social platforms, stochastic solvers are the most suitable and especially less time consuming than approximation methods.

English abstract

Topic modeling is one of the most widely used NLP techniques for discovering hidden patterns and latent relationships in text documents. Latent Dirichlet Allocation (LDA) is one of the most popular methods in this field. There are different approaches to tuning the parameters of LDA models such as Gibbs sampling, variational method, or expected propagation. This paper aims to compare individual LDA parameter tuning approaches with respect to their performance and accuracy on textual data from the financial domain. From our results, it can be concluded that for text datasets obtained from financial social platforms, stochastic solvers are the most suitable and especially less time consuming than approximation methods.

Keywords

Financial Messages; Latent Dirichlet Allocation; LDA; NLP; Text analysis; Topic Modeling

Key words in English

Financial Messages; Latent Dirichlet Allocation; LDA; NLP; Text analysis; Topic Modeling

Authors

JANKOVÁ, Z.

RIV year

2024

Released

01.04.2023

Publisher

Bucharest University of Economic Studies

Location

Bucharest, Romania

ISBN

0424-267X

Periodical

Economic Computation and Economic Cybernetics Studies and Research

Volume

57

Number

1

State

Romania

Pages from

267

Pages to

282

Pages count

16

URL

https://ecocyb.ase.ro/nr2023_1/2023_1_17_ZuzanaJANKOVA_online.pdf

BibTex

@article{BUT183297,
  author="Zuzana {Janková}",
  title="Latent Dirichlet Allocation (LDA) Approximation Analysis of Financial-Related Text Messages",
  journal="Economic Computation and Economic Cybernetics Studies and Research",
  year="2023",
  volume="57",
  number="1",
  pages="267--282",
  doi="10.24818/18423264/57.1.23.17",
  issn="0424-267X",
  url="https://ecocyb.ase.ro/nr2023_1/2023_1_17_ZuzanaJANKOVA_online.pdf"
}