Topmodpy: A Simple Python Script for Topic Modeling

Musunuru, Kamakshaiah

doi:10.2139/ssrn.3746196

Cited by 1 publication

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Early topic modeling was first described in 1998 by Papadimitriou, Raghavan, Tamaki, and Vempala [18]. In 1999, Thomas Hofmann introduced Probabilistic Latent Semantic Analysis (PLSA) [19]. Latent Dirichlet Allocation (LDA), perhaps the most widely used topic model today, is a generalization of PLSA [20].…”

Section: Topic Modeling and Interpretationmentioning

confidence: 99%

Analyzing ESG News Articles and Research Papers through LDA (Latent Dirichlet Allocation)

Bowen,

Min

2023

APJCRI

View full text Add to dashboard Cite

The evolution of information technology has led to a surge in the publication of research papers across diverse fields. This surge has presented a challenge for academics in finding relevant literature and unexplored research areas. To address this, the current study designs a research gap finding framework using Topic Modeling methodology. For this study, 4,596 news articles were collected from reputable Korean sources such as Chosun, JoongAng, Hankyoreh, and Kyunghyang, alongside 456 ESG-related academic research papers. The Topic Modeling method of Latent Dirichlet Allocation (LDA) was utilized to distill subtopics from these resources. The use of LDA in this context proved crucial as it allowed for the extraction of nuanced subtopics from a large dataset, aiding the identification of underexplored themes. The comparative analysis of subtopics extracted from news articles and academic papers demonstrated a thematic consistency, but a distinct difference in specific subtopics. Results of the study indicated that academic research tends to overlook topics such as community cooperation and industry-academic partnerships. Furthermore, social issues like child labor, forced labor, and workplace health and safety received relatively minimal attention in academic papers. In-depth examination of these results revealed a significant gap between popular social issues and existing academic research, implying a need for more focused research in these areas. This study significantly aids academics in better comprehending current social issues and identifying research gaps. The proposed framework, leveraging both news articles and academic papers, provides a practical tool for academics to stay current with societal issues and recognize the areas lacking in academic exploration. The framework's practicality represents a vital advancement in the field, offering scholars a robust tool for enriching academic literature and engaging more effectively in societal discourse. This dual utility not only bridges the gap between theoretical research and real-world applications but also fosters a more informed and nuanced understanding of the subject matter, thereby contributing significantly to both scholarly inquiry and public conversation.

show abstract

Section: Topic Modeling and Interpretationmentioning

confidence: 99%