Topic-Noise Models: Modeling Topic and Noise Distributions in Social Media Post Collections

Churchill, Rob; Singh, Lisa

doi:10.1109/icdm51629.2021.00017

Cited by 7 publications

(4 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different strategies for mitigating noise have been proposed, including (1) modelling documents as a graph or network and removing noisy parts of the network before detecting topics and (2) adjusting generative probabilistic models to account for short, noisy text. A new approach that focuses on building topic‐noise models by approximating both the underlying topic and noise probability distributions is an important step for identifying topics in short, noisy posts (Churchill & Singh, 2021).…”

Section: Constructing Variables From Big Datamentioning

confidence: 99%

“…Determining the topic(s) in an article or post is central to understanding how the dynamics of conversation about different topics are changing through time. While different approaches have been proposed for extracting topics 7. from the text (see surveys for more detail (Qiang et al, 2020;Churchill & Singh, in press), new methods that adequately handle the noise of social media text streams are still in their infancy (Churchill & Singh, 2020Dieng et al, 2019;Wang et al, 2018, Qiang et al, 2017. Different strategies for mitigating noise have been proposed, including (1) modelling documents as a graph or network and removing noisy parts of the network before detecting topics and (2) adjusting generative probabilistic models to account for short, noisy text.…”

Section: Cons Truc Ting Variab Le S From B Ig Datamentioning

confidence: 99%

“…A new approach that focuses on building topic-noise models by approximating both the underlying topic and noise probability distributions is an important step for identifying topics in short, noisy posts (Churchill & Singh, 2021).…”

Section: Cons Truc Ting Variab Le S From B Ig Datamentioning

confidence: 99%

See 2 more Smart Citations

Environmental change and human mobility: Opportunities and challenges of big data

Martin

Singh

2022

International Migration

Self Cite

View full text Add to dashboard Cite

This paper explores the use of big data in understanding population movements in the context of environmental change, with particular attention to publicly available information from social media and newspapers. After discussing causal factors, the paper discusses the benefits and challenges of using big data, and mathematical models that may help capture the complexity of human mobility in the context of environmental change. We conclude that big data can be an important part of a toolkit in discerning drivers of migration and displacement. It is no panacea, however; the paper discusses potential problems in its use, particularly if big data are analysed to prevent people from moving out of harm's way. Therefore, partnerships between computer scientists and subject matter experts and practitioners are essential to ensuring that what is gathered from big data is usable, credible, and protects the privacy and rights of those who may be displaced.

show abstract

Section: Constructing Variables From Big Datamentioning

confidence: 99%

Section: Cons Truc Ting Variab Le S From B Ig Datamentioning

confidence: 99%

See 1 more Smart Citation

Environmental change and human mobility: Opportunities and challenges of big data

Martin

Singh

2022

International Migration

Self Cite

View full text Add to dashboard Cite

show abstract

“…This contrasts to our approach of using an unsupervised algorithm to generate comparisons between multiple topic models, although we do impose a binary classification of topics related to semantic vs logistical content. Churchill & Singh [31] consider topic models from the perspective of topic-noise, using pre-trained model in an ensemble with LDA to generate more diverse and coherent topics. This approach also uses the semantic information from word emebeddings but is specifically tuned to use topicnoise as a discriminator that can be used either during or after topic generation.…”

Section: Related Workmentioning

confidence: 99%

Leveraging Multiple Representations of Topic Models for Knowledge Discovery

2022

View full text Add to dashboard Cite

Topic models are often useful in categorization of related documents in information retrieval and knowledge discovery systems, especially for large datasets. Interpreting the output of these models remains an ongoing challenge for the research community. The typical practice in the application of topic models is to tune the parameters of a chosen model for a target dataset and select the model with the best output based on a given metric. We present a novel perspective on topic analysis by presenting a process for combining output from multiple models with different theoretical underpinnings. We show that this results in our ability to tackle novel tasks such as semantic characterization of content that cannot be carried out by using single models. One example task is to characterize the differences between topics or documents in terms of their purpose and also importance with respect to the underlying output of the discovery algorithm. To show the potential benefit of leveraging multiple models we present an algorithm to map the term-space of Latent Dirichlet Allocation (LDA) to the neural document-embedding space of doc2vec. We also show that by utilizing both models in parallel and analyzing the resulting document distributions using the Normalized Pointwise Mutual Information (NPMI) metric we can gain insight into the purpose and importance of topics across models. This approach moves beyond topic identification to a richer characterization of the information and provides a better understanding of the complex relationships between these typically competing techniques.

show abstract

Dynamic Topic-Noise Models for Social Media

Churchill

Singh

2022

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

Topic-Noise Models: Modeling Topic and Noise Distributions in Social Media Post Collections

Cited by 7 publications

References 19 publications

Environmental change and human mobility: Opportunities and challenges of big data

Environmental change and human mobility: Opportunities and challenges of big data

Leveraging Multiple Representations of Topic Models for Knowledge Discovery

Dynamic Topic-Noise Models for Social Media

Contact Info

Product

Resources

About