方法对比

	Network Text Analysis	Topic Modeling for Communication Research
领域	Communication	Communication
方法族	Process / pipeline	Machine learning
起源年份	2002	2003
提出者	Corman et al. (centering resonance analysis); network text tradition	David Blei et al. (LDA); Roberts, Stewart & Tingley (STM)
类型	Representation and analysis of text as networks of linked concepts	Unsupervised probabilistic model of latent themes in document collections
开创性文献	Corman, S. R., Kuhn, T., McPhee, R. D., & Dooley, K. J. (2002). Studying complex discursive systems: Centering resonance analysis of communication. Human Communication Research, 28(2), 157–206.	Blei, D. M., Ng, A. Y., & Jordan, M. I. (2003). Latent Dirichlet allocation. Journal of Machine Learning Research, 3, 993–1022.
别名	Text network analysis, Centering resonance analysis, Concept network analysis, Ağ Tabanlı Metin Analizi	LDA for communication, Structural topic modeling in communication, Topic models for media texts, İletişim Araştırmaları için Konu Modelleme
相关	4	3
摘要	Network text analysis represents the content of text not as counts of words or topics but as a network of concepts linked by their relationships, then applies social-network methods to reveal which ideas are central and how they connect. Centering resonance analysis (CRA), introduced by Corman and colleagues in 2002, is a leading variant that builds concept networks from the noun phrases that structure discourse.	Topic modeling is an unsupervised technique for discovering the latent themes that run through a large collection of documents, representing each document as a mixture of topics and each topic as a distribution over words. In communication research it surfaces the issues, frames, and themes in news archives, social media, and political text at a scale no manual reading can match, with Latent Dirichlet Allocation (LDA) and the Structural Topic Model (STM) as the dominant variants.