Methodology

Optimal text-based time-series indices

Choosing the aggregation scheme by optimising it against the target, instead of fixing it in advance.

The R package sentometrics to compute, aggregate, and predict with textual sentiment

The vignette for the sentometrics R package — one coherent workflow from raw texts to a sentiment-based forecast.

Sentiment and Econometrics: Toward a Unified Framework of Textual Sentiment Analysis for Economic and Financial Applications

The doctoral work behind the sentometrics framework — a unified treatment of textual sentiment analysis for economic and financial applications.

Econometrics meets sentiment: An overview of methodology and applications

The survey that defines the field: how qualitative sentiment in text becomes a quantitative variable, and how that variable enters an econometric analysis.

Questioning the news about economic growth: Sparse forecasting using thousands of news-based sentiment values

The methodology behind the sentometrics package, applied to US economic growth — and the source of the topical US sentiment indices.

Modeling Latent Variables in Economics and Finance

Doctoral work on latent-variable modelling in economics and finance, and one of the foundations of the sentometrics programme.