An Enhanced BERTopic Framework and Algorithm for Improving Topic Coherence and Diversity

Sawant, Sahil; Yu, Jinhong; Pandya, Kirtikumar; Ngan, Chun-Kit; Bardeli, Rolf

doi:10.1109/hpcc-dss-smartcity-dependsys57074.2022.00332

Cited by 8 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 5 shows the results of performance metrics of the LDA model, BERTopic model with MMR, OpenAI, and keyBERT models for topic modeling. In the study, evaluation metrics commonly used in the literature were selected for topic modeling [ 51 , 52 ]. These metrics: Topic Coherence, Normalized Pointwise Mutual Information (NPMI), (C_v) and Topic Diversity.…”

Section: Resultsmentioning

confidence: 99%

Exploring climate change discourse on social media and blogs using a topic modeling analysis

Gokcimen,

Das

2024

Heliyon

View full text Add to dashboard Cite

Section: Resultsmentioning

confidence: 99%

Exploring climate change discourse on social media and blogs using a topic modeling analysis

Gokcimen,

Das

2024

Heliyon

View full text Add to dashboard Cite

“…For classification, we utilized BERT, specifically BERTopic [21][22][23], for topic definition and KeyBERT [24] for the keyword extraction from documents. Guided topic modeling is enhanced through similarity calculations, employing methods like Spacy cosine similarity, Scikit-learn cosine similarity, and Jaccard similarity [25][26][27].…”

Section: Methodsmentioning

confidence: 99%

Discovering Trends in the Digitalization of Shipping: An Exploratory Study into Trends Using Natural Language Processing

Aerts,

Mathys

2024

JMSE

View full text Add to dashboard Cite

This study investigates digitalization in the shipping industry by analyzing over 500 industry presentations from an eight-year span to discern key trends and nascent signals. Employing optical character recognition, advanced natural language processing techniques, and similarity metrics, the research enhances topic interpretability. Through Theil–Sen regressions and diffusion metrics, it identifies trends and emerging signals, noting a rise in interest in smart ports and supply chain management, signaling a shift toward more intelligent technology integration. However, attention to supply chain management shows a decline. The research tracks a shift from broad technology themes to specific areas like cybersecurity and blockchain, reflecting a narrative pivot to tackle particular digital challenges and opportunities. The study detects weak signals, including terms like “subsea” and “drone”, suggesting forthcoming industry innovations and shifts, notably toward ESG considerations. An additional machine learning analysis corroborates findings on key topics like energy efficiency and crew welfare, also spotlighting virtual disaster recovery and ERP projects as emerging areas of interest. This work aids in comprehending the fluid digitalization landscape in shipping, highlighting the sector’s ongoing evolution, and underscoring the need for further inquiry into autonomous shipping and related domains.

show abstract

Noise-aware celestial clustering for hot topic detection from microblog datasets with not well-separated topics

kumar,

Samuel

2024

Knowl Inf Syst

View full text Add to dashboard Cite

An Enhanced BERTopic Framework and Algorithm for Improving Topic Coherence and Diversity

Cited by 8 publications

References 4 publications

Exploring climate change discourse on social media and blogs using a topic modeling analysis

Exploring climate change discourse on social media and blogs using a topic modeling analysis

Discovering Trends in the Digitalization of Shipping: An Exploratory Study into Trends Using Natural Language Processing

Noise-aware celestial clustering for hot topic detection from microblog datasets with not well-separated topics

Contact Info

Product

Resources

About