What are the three NLP tasks integrated in Garcia Quevedo, Glaser, and Verzat's method for relevant data selection, and what is the purpose of each?

The method integrates sentiment analysis to classify expressions by affective tenor, topic modeling to identify underlying data patterns and clustering, and information retrieval to find similarities at the sentence and paragraph levels. These tasks together support exploration, understanding, and relevant data selection before manual inductive analysis.

Garcia Quevedo, Glaser, and Verzat's relevant data selection method combines three NLP tasks via LLM-based algorithms. Sentiment analysis classifies expressions according to an affective tenor, meaning it sorts data by emotional or evaluative tone. Topic modeling identifies underlying data patterns and performs clustering, which helps reveal themes or groupings across the dataset. Information retrieval finds similarities at the sentence and paragraph levels, enabling targeted search through the content. This combined approach diversifies data exploration and enables triangulation during selection. The authors recommend using these tasks before manual inductive analysis so that the entire dataset can be explored and understood, and a focused selection can be created with conscious, step-by-step choices aligned with the research question.