手法を比較

	Topic Modeling for Communication Research	Automated Content Analysis
分野	Communication	Communication
系統	Machine learning	Process / pipeline
提唱年	2003	2013
提唱者	David Blei et al. (LDA); Roberts, Stewart & Tingley (STM)	Justin Grimmer & Brandon Stewart (synthesis)
種類	Unsupervised probabilistic model of latent themes in document collections	Computational pipeline for measuring features of large text corpora
原典	Blei, D. M., Ng, A. Y., & Jordan, M. I. (2003). Latent Dirichlet allocation. Journal of Machine Learning Research, 3, 993–1022.	Grimmer, J., & Stewart, B. M. (2013). Text as data: The promise and pitfalls of automatic content analysis methods for political texts. Political Analysis, 21(3), 267–297.
別名	LDA for communication, Structural topic modeling in communication, Topic models for media texts, İletişim Araştırmaları için Konu Modelleme	Computational content analysis, Text-as-data analysis, Automated text analysis, Otomatik İçerik Analizi
関連	3	4
概要	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.	Automated content analysis is the computational measurement of text features at a scale impossible by hand, using natural-language processing and machine learning to classify, scale, or discover the content of large corpora. Synthesized for the social sciences by Grimmer and Stewart's 2013 'Text as Data,' it spans supervised classification, unsupervised discovery, and scaling, all unified by the principle that automated methods augment but do not replace careful human judgment and validation.