Quantitative analysis of large amounts of journalistic texts using topic modelling

Jacobi, Carina; Atteveldt, Wouter van; Welbers, Kasper

doi:10.1080/21670811.2015.1093271

Cited by 383 publications

(255 citation statements)

References 13 publications

Supporting

Mentioning

247

Contrasting

Unclassified

Order By: Relevance

“…Besides the analysis of open-ended survey questions on climate change (Tvinnereim and Fløttum 2015), STM has so far been applied to the analysis of documents produced by organizations opposed to action on climate change (Farrell 2016), classification of Arab Muslim cleric writings, and reactions to the Edward Snowden case (Lucas et al 2015). Related topic modeling methods have been employed to classify think tank statements on climate science and policy (Boussalis and Coan 2016) as well as analyze trends in newspaper coverage of nuclear technology over time (Jacobi et al 2016),…”

Section: Methodsmentioning

confidence: 99%

Public perceptions of air pollution and climate change: different manifestations, similar causes, and concerns

2016

View full text Add to dashboard Cite

Co-benefits are seen as a key factor for overcoming the problems of collective action and extended time horizons holding back mitigation of global warming. The mechanism behind this hypothesis is that public acceptance of mitigation policies constitutes a crucial limiting factor, necessitating ancillary gains such as clean air for mitigation policy to be politically robust. However, the public's preference for local pollution mitigation and concomitant failure to appreciate the benefits of global warming mitigation is assumed rather than demonstrated. In this paper, we show, first, that people distinguish between the physical manifestations of air pollution and global warming, and second, that they see both phenomena as arising from the same causes as well as having negative impacts on humans. Specifically, using a survey experimental design with open-ended questions in an urban Chinese setting, we demonstrate that citizens relate glacier melt and sea-level rise to global warming, while linking the local phenomenon of smog almost exclusively to air pollution. At the same time, respondents link impacts on humans and vehicle/industrial pollution topics with both air pollution and global warming. These findings are relevant to decision-makers as they suggest that the public values mitigation of global warming in its own right. Our novel method may shed new light on a range of issues relating to energy, the economy, and environmental issues.

show abstract

Section: Methodsmentioning

confidence: 99%

Public perceptions of air pollution and climate change: different manifestations, similar causes, and concerns

2016

View full text Add to dashboard Cite

show abstract

“…It is a key task in topic modeling to determine the number of examined topics (Gläser et al 2017;Jacobi et al 2016). In this study, the authors sought to identify the maximum number of topics to describe the data while still being able to explain and distinguish between the topics well.…”

Section: Topic Modeling As a Methodsology To Approach Citizen Visionsmentioning

confidence: 99%

“…If the results of the two analyses conform, topic modeling could be seen as a viable alternative for deliberative analysis (cf. Gläser et al 2017;Jacobi et al 2016). Differences, on the other hand, would highlight the particular character of each methodology.…”

Section: Findings From the Two Methodsologiesmentioning

confidence: 99%

“…Thereby, this paper also contributes to issues relating to internal validity and external generalizability of topic modeling, which are still under discussion. This follows the suggested procedure of evaluating validity by comparing results from machine coding with those from manual coding when making sense of texts (Jacobi, van Atteveldt, and Welbers 2016;Gläser et al 2017). Furthermore, the paper provides a methodological comparison of these two approaches.…”

Section: Europe Seeks Citizen Visions For Future Research and Innovatmentioning

confidence: 99%

See 1 more Smart Citation

Sustainable Futures: Comparing Methodologies for Analyzing Citizen Visions in Europe

Repo

Matschoss

Timonen

2017

View full text Add to dashboard Cite

There are increasing calls for engaging citizens in the development of future outlooks. At the same time, large-scale public engagement activities warrant appropriate methods for analyzing their outcomes. This paper reviews how topic modeling could provide such a methodology, which both accounts for all textual data collected in public engagement activities, however large in scope, yet also allows for meaningful topical analysis. It compares topic modeling results concerning a corpus of 179 citizen visions from 30 European countries on desirable and sustainable futures to those acquired through deliberative analysis. While both methodologies contend that European citizens' outlook consists of education, sustainability in the economy, health concerns, and fairness in communities, and the particular strengths of topic modeling relate to its documentability, repeatability, cost efficiency, and scalability. Topic modeling can also be considered to support public engagement analytically from the perspective of knowledge formation rather than that of common sense.

show abstract

“…They used a relatively recent and innovative technique, Structural Topic Modelling, which allows to detect latent themes in the statements. Topic modelling approaches were also used to analyse themes in large journalistic coverage corpora [Jacobi, Atteveldt and Welbers, 2016].…”

Section: Computer-aided Analyses In Public Communication Of Science Amentioning

confidence: 99%

Computer-aided text analysis: an open-aired laboratory for social sciences

Castelfranchi¹

2017

JCOM

View full text Add to dashboard Cite

Thanks, on the one hand, to the extraordinary availability of colossal textual archives and, on the other hand, to advances in computational possibilities, today the social scientist has at their disposal an extraordinary laboratory, made of millions of interacting subjects and billions of texts. An unprecedented, yet challenging, opportunity for science. How to test, corroborate models? How to control, interpret and validate Big Data? What is the role of theory in the universe of patterns and statistical correlations? In this article, we will show some general characteristics of the use of computational tools for the analysis of texts, and some applications in the areas of public communication of S&T and Science and Technology Studies (STS), also showing some of their limitations and pitfalls. AbstractPublic understanding of science and technology; Representations of science and technology; Science communication: theory and models KeywordsThe information deluge has flooded social sciences. Governments around the world digitalize and make available laws and projects, parliamentary debates and questions, social and economic indicators, party funding, government programs, polls and public opinion polls, maps, and models, due to new governance practices (and the new rhetoric) of transparency, accountability, participation and deliberation. Dynamics of globalized markets and customer demand force large and small media outlets (newspapers and magazines, TVs and radios, network on the Internet and on social networks) to make it possible to download huge databases, including archives of what was written, said, sung over a period of decades.Publishers publish their books in digital format. Libraries and newspaper libraries, as well as corporations such as Google, file millions of publications containing a considerable amount of literature and world history. New forms of sociability, on and offline, new modes, effects and affections of contemporary individualism lead tens of millions of users to illustrate, narrate, edit, and caption on web 2.0 their tastes and intimacies, their daily life and their memories, the interactions, associations, conflicts, dreams, resentments and confessions. We are exposed to millions of gigabytes of information and, at the same time, we are tireless

show abstract

Quantitative analysis of large amounts of journalistic texts using topic modelling

Cited by 383 publications

References 13 publications

Public perceptions of air pollution and climate change: different manifestations, similar causes, and concerns

Public perceptions of air pollution and climate change: different manifestations, similar causes, and concerns

Sustainable Futures: Comparing Methodologies for Analyzing Citizen Visions in Europe

Computer-aided text analysis: an open-aired laboratory for social sciences

Contact Info

Product

Resources

About