What topic modeling could reveal about the evolution of economics

Abstract

The paper presents the topic modeling technique known as Latent Dirichlet Allocation (LDA), a form of text-mining aiming at discovering the hidden (latent) thematic structure in large archives of documents. By applying LDA to the full text of the economics articles stored in the JSTOR database, we show how to construct a map of the discipline over time, and illustrate the potentialities of the technique for the study of the shifting structure of economics in a time of (possible) fragmentation.

Classifying economics for the common good: a note on the links between sustainable development goals and JEL codes

“…This would require in-depth bibliometric, topic modelling, and science mapping analyses (cf. Aria and Cuccurullo, 2017; Körfgen et al , 2018; Ambrosino et al , 2019; LaFleur, 2019). These are left for future studies.…”

Start-ups survival through a crisis. Combining machine learning with econometrics to measure innovation

“…1 A second approach, based on machine learning, takes the form of unsupervised algorithms which create a data partition without any a priori restriction on the number and type of categories to be generated. Clustering algorithms (Macqueen 1967), self-organising maps (Carlei and Nuccio 2014) and, more recently, topic modelling for text analysis of the economic literature (Ambrosino et al 2018) belong to this group. In unsupervised algorithms, the model validation is pursued by an expost educated interpretation of the result.…”

“…We claim that research in economics can take advantage of the latter approaches. Even though someone has envisioned the end of economic theory and the transition towards a purely datadriven type of science (Anderson 2008;Prensky 2009), other authors suggest that the large availability of data, which reveals the complexity of the relationships in the observed reality, actually calls for more theory (Kitchin 2014;Ambrosino et al 2018;Nuccio and Guerzoni 2019;Carota, Durio, and Guerzoni 2014;Gould 1981). Data and its analysis can still act as a powerful hypothesis-mining engine (Jordan 1998;Carota, Durio, and Guerzoni 2014) and provide new theoretical ideas, which then need to be filtered through a theoretical framework.…”

Research Community Analytic Tool Based on Topic Modeling and Network Analysis

“…Кроме патентов, ММТ широко использовался для других типов текстовых данных. Многие исследования изучали научную литературу, опубликованную в разных реферируемых журналах [2,8,9], или все экономические статьи из базы данных вроде JSTOR [10]. Более того, некоторые исследования специально фокусировались на литературе, посвященной таким темам, как информационная безопасность [11] или биоинформатика [12].…”

“…Более того, некоторые исследования специально фокусировались на литературе, посвященной таким темам, как информационная безопасность [11] или биоинформатика [12]. Тогда как одни ученые смотрят только на аннотации статей [2], другие изучают полные тексты статей [10].…”