Text as Data

2019/09/01 by Matthew Gentzkow, Bryan Kelly, Matt Taddy · 75 citations
Social Sciences · Computer Science · #Media Influence and Politics #Advanced Text Analysis Techniques #Computational and Text Analysis Methods

paper · doi:10.1257/jel.20181020

Abstract

An ever-increasing share of human interaction, communication, and culture is recorded as digital text. We provide an introduction to the use of text as an input to economic research. We discuss the features that make text different from other forms of data, offer a practical overview of relevant statistical methods, and survey a variety of applications. (JEL C38, C55, L82, Z13)

Citations

• The horseshoe estimator for sparse signals
• Probabilistic topic models
• What Drives Media Slant? Evidence From U.S. Daily Newspapers

Cited by

• Generative AI for Economic Research: Use Cases and Implications for Economists
• Enlightenment Ideals and Belief in Progress in the Run-up to the Industrial Revolution: A Textual Analysis
• Data and the Aggregate Economy
• What is Certain about Uncertainty?
• Deep Learning for Economists
• Computational Text Analysis for Building and Testing Social Theory
• The Global Impact of Brexit Uncertainty
• Measuring issue salience for political parties using LLMs
• The Diffusion of New Technologies
• Beyond standardization: a comprehensive review of topic modeling validation methods for computational social science research
• New Frontiers: The Origins and Content of New Work, 1940–2018
• Review bombing: ideology-driven polarisation in online ratings: The case study of The Last of Us (part II)
• Officer-Involved: The Media Language of Police Killings
• DECODEM: Data Extraction from Corporate Organizational Documents via Enhanced Methods
• Scaling Point-in-Time Language Models
• The effect of publicly co-funded industry-science collaboration on scientific production
• Feeling the heartbeat of regions: local news and economic sentiments
• “The risks cannot be compensated”: The willingness to donate DNA for science and its relationship with economic preferences
• Narrative forecasts
• Why Do You Want to Be a Teacher? A Natural Language Processing Approach
• Large Language Models have Chain-of-Affect
• Narratives to Numbers: Large Language Models and Economic Policy Uncertainty
• Forecasting with Economic News
• Semantic maps and metrics for science Semantic maps and metrics for science using deep transformer encoders
• Words Matter: Forecasting Economic Downside Risks with Corporate Textual Data
• Making Interpretable Discoveries from Unstructured Data: A High-Dimensional Multiple Hypothesis Testing Approach
• Computational timeline reconstruction of the stories surrounding Trump: Story turbulence, narrative control, and collective chronopathy
• The sin of words: Censorship and self‐censorship in China during the Qing dynasty (1644–1911)
• FinBERT: A Large Language Model for Extracting Information from Financial Text*
• Primer for Experimental Methods in Organization Theory
• Artificial intelligence as a general-purpose technology: an historical perspective
• Fundraising, Governance and Environmental Ethics: Evidence from Equity Crowdfunding
• Diverting Domestic Turmoil
• Political Rhetoric and Racial Discrimination in Arrests for Drugs
• A Quantitative Approach to Multimodality: An application to the creative process of Van Gogh
• Tracking Climate and Environmental Attention: A News‐Based Composite Index
• Firm‐Level Climate Change Exposure
• Complex business ecosystem intelligence using AI-powered visual analytics
• Dirty talk: Media discourse and the struggle over South Africa’s coal transition
• Geopolitics, Geoeconomics and Risk: A Machine Learning Approach
• The impact of the FinTech revolution on the future of banking: Opportunities and risks
• Framing contestation and public influence on policymakers: evidence from US artificial intelligence policy discourse
• Whatever it takes to understand a central banker — Embedding their words using neural networks
• Remembering Li Keqiang: Policy Divergence in Zhongnanhai and Its Economic Consequences
• AWARE-FX: An Auditable Knowledge-Guided AI System for Measuring Corporate Foreign-Exchange Hedging Disclosure
• Signalling sustainability under uncertainty: how institutionalized digital platforms shape corporate green transformation in China
• Overcoming the innovation threshold through innovative public procurement: evidence from CERN
• Measuring Corporate Culture Using Machine Learning
• Hamiltonian Monte Carlo for Regression with High-Dimensional Categorical\n Data
• News media versus FRED‐MD for macroeconomic forecasting
• A Financial Brain Scan of the LLM
• Forecasting with Economic News
• Measuring news sentiment
• Large Language Models for Statistical Inference: Context Augmentation with Applications to the Two-Sample Problem and Regression
• Position: Use Sparse Autoencoders to Discover Unknowns
• Categorizing nonprofit organizations according to their field of activity: A discussion of rule-based categorization and machine learning, and recommendations for implementation
• The effect of central bank communication on sovereign bond yields: The case of Hungary. [europepmc]
• Innovation indicators based on firm websites-Which website characteristics predict firm-level innovation activity? [europepmc]
• Health by Words-A Content Analysis of Political Manifestos in the Portuguese Elections of 1975. [europepmc]
• Computational timeline reconstruction of the stories surrounding Trump: Story turbulence, narrative control, and collective chronopathy. [europepmc]
• Multifractal analysis of social media use in financial markets. [europepmc]
• China's environmental policy intensity for 1978-2019. [europepmc]
• The semi-automatic classification of an open-ended question on panel survey motivation and its application in attrition analysis. [europepmc]
• Schools' We-mentality and Students' Civic Engagement - A Text-based Approach. [europepmc]
• Two is better than one: Using a single emotion lexicon can lead to unreliable conclusions. [europepmc]
• Citizens' economic recovery models for a pandemic. [europepmc]
• Quality indices for topic model selection and evaluation: a literature review and case study. [europepmc]
• Towards understanding policy design through text-as-data approaches: The policy design annotations (POLIANNA) dataset. [europepmc]
• Can digital transformation change a firm's green innovation strategy? Evidence from China's heavily polluting industries. [europepmc]
• Does updating education curricula accelerate technology adoption in the workplace? Evidence from dual vocational education and training curricula in Switzerland. [europepmc]
• Word embedding for social sciences: an interdisciplinary survey. [europepmc]
• A geolocated dataset of German news articles. [europepmc]
• Artificial intelligence adoption and corporate ESG performance: evidence from a refined large language model. [europepmc]
• LLMs outperform outsourced human coders on complex textual analysis. [europepmc]
• Hybrid Analysis of Videoconference Technology Use by Aging-in-Place Organizations to Promote Social Engagement for Older Adults: A Scoping Review with Latent Topic Modeling. [europepmc]
• Radiomic features and carotid stenosis in periodontitis a two stage bootstrap and multimodal machine learning study. [europepmc]

Related

• Text as Data: Real-time Measurement of Economic Welfare
• Opening Up to Big Data : Computer-Assisted Analysis of Textual Data in Social Sciences
• Data and the Aggregate Economy
• Big data in economics
• Text as Behavior