Latent Dirichlet Allocation

Campbell, Joshua Charles; Hindle, Abram; Stroulia, Eleni

doi:10.1016/b978-0-12-411519-4.00006-9

Cited by 220 publications

(84 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…'ϴ' is the topic distribution for document i. 'α', 'β' are the hyperparameters of Dirichlet distribution (Source: (Blei, Ng, & Jordan, 2003)).…”

Section: Topicsmentioning

confidence: 99%

“…These topics remain latent and will only emerge during the modelling process. It is based on the basic idea that each document can be expressed as a distribution of topics and each topic can be described by a distribution of words (Blei et al, 2003). The basic terminology used in LDA is based on the language of 'text collection', referring to entities such as "words", "documents" and "corpora".…”

Section: A2 Mathematical Background: Ldamentioning

confidence: 99%

“…The basic terminology used in LDA is based on the language of 'text collection', referring to entities such as "words", "documents" and "corpora". These terms are defined as (Blei et al, 2003):…”

Section: A2 Mathematical Background: Ldamentioning

confidence: 99%

“…The basic idea is that documents are represented as random mixtures over latent topics, where each topic is characterized by a distribution over words. For the list of assumptions adopted in the basic LDA model, please refer to the Section 3 of Blei et al, (2003). The probability distribution function of a k-dimensional Dirichlet random variable θ is illustrated in eq (1), where θ can take values in the (k-1)-simplex (a k-vector θ lies in the (k-1)-simplex if ≥ 0, ∑ = 1) ,…”

Section: Appendixmentioning

confidence: 99%

See 3 more Smart Citations

Energy justice, built environment and poverty: A deep-narrative analysis of energy cultures in Brazil, India and Nigeria for policy modelling

Debnath¹,

Bardhan²,

Darby³

et al. 2020

Preprint

View full text Add to dashboard Cite

Distributive energy justice as a policy instrument is crucial to meet UN-SDG 7 targets at the poorest section of rapidly urbanising Global South. This study mobilised this policy instrument to derive poverty alleviation focal points across slum rehabilitation housing in Brazil, India and Nigeria. It emphasises on the social dimensions of energy services which currently remains a literature gap. The core methodological innovation lies in the nested deep narrative analysis approach using topic modelling and grounded theory, developed in Debnath et. al. (2020). This paper presents the application of this methodology through the analysis of energy culture and built environment in poverty. Narratives of poverty and energy services were collected through focus group discussions that served as the primary data object. Topic modelling was performed to remove directionality biases in the narratives using a Latent Dirichlet Allocation (LDA)-based algorithm, which was then strengthen using grounded theory. Cross-validation of the topic models was performed using science-mapping of state-of-the-art literature in ‘energy justice’ and ‘poverty’ with an LDA-tuning algorithm. Results converged around seven key topics that probabilistically highlighted the grounded reality of complex interactions between poverty, built environment and human-scale energy services. The derived policy-focal points illustrated that built environment design, land use planning and governance are crucial to the success of SDG-7 in poverty. It showed how utilities and distribution companies can strengthen their energy governance to support social policies and improve the quality and reliability of electric power. The high diffusion of ICTs in the study area can act as a platform of change to bring distributive justice.

show abstract

“…'ϴ' is the topic distribution for document i. 'α', 'β' are the hyperparameters of Dirichlet distribution (Source: (Blei, Ng, & Jordan, 2003)).…”

Section: Topicsmentioning

confidence: 99%

Section: A2 Mathematical Background: Ldamentioning

confidence: 99%

Section: A2 Mathematical Background: Ldamentioning

confidence: 99%

Section: Appendixmentioning

confidence: 99%

See 2 more Smart Citations

Energy justice, built environment and poverty: A deep-narrative analysis of energy cultures in Brazil, India and Nigeria for policy modelling

Debnath¹,

Bardhan²,

Darby³

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…The matrix element contains the frequency of the i term in the nth document. produce a weighted list of topics for each document (Campbell, 2014). The distribution used to get the distribution of topics per document is called Dirichlet distribution, then in the generative process for LDA, the results from Dirichlet are used to allocate words to documents for different topics.…”

Section: Transformation Of Title and Abstractmentioning

confidence: 99%

Using LDA for Innovation Topic of Technology : Quantum Dots Patent Analysis

Purba¹,

Nooraeni²

2020

Proceedings of the Proceedings of the 1st International Conference on Statistics and Analytics, ICSA 2019, 2-3 August 2019, Bog

View full text Add to dashboard Cite

This study seeks to explore information about one of nanotechnology, quantum dots (QDs), through analysis of patent information. QDs patent documents obtained from the United States international patent database, the USPTO, use web scraping. In total, 3914 patents from 1988 to 2016 were taken and archived for analysis. This paper discusses how to apply Latent Dirichlet Allocation (LDA), a topic model, in a trend analysis methodology that exploits patent information. After the text preprocessing and transformation, the number of topics is decided using the log likelihood value. Then LDA model is used for identifying underlying topic structures based on latent relationships of technological words extracted. We extracted words from 6 relevant topics and showed that these topics are highly meaningful in explaining technology applications of QDs.

show abstract