NaNaLe

Online discourses on social media, blogs, and review platforms have become central spaces where consumers discuss, assess, and share experiences about food. Enormous volumes of text data accumulate, and can be systematically analyzed to identify reliable patterns and tracked over extended periods. The texts are often short, fragmentary, and highly context-dependent; diverse languages, slang, emojis, and hashtags complicate processing and analysis.

Despite this heterogeneity, these data provide valuable insights into attitudes, values, and expectations around sustainability topics such as organic products, animal welfare, regional origin, packaging, and the avoidance of food waste. The challenge lies in identifying patterns in large text collections that enable a systematic, theory-guided classification into narratives without losing fine semantic nuances.

The goal is to identify dominant topics in the extensive text material and to show how attitudes and emotional states change over time. Beyond mere topic distributions, tone, audience positions, and lines of conflict can be analysed: who argues for what (e.g., price versus sustainability), which roles are occupied in narratives (retailers, consumers, farmers), and which solutions are proposed. Network analyses enable the linking of terms and ideas. The results provide indications for effective online communication strategies on sustainability on the part of grocery retailers.

The approach is based on a multi-stage text analysis that combines modern AI-assisted methods with classical text-analytic techniques. First, the texts undergo standardized preprocessing (language identification, normalization, lemmatization, handling of emojis/hashtags). Second, a central component is an AI-supported, topic-oriented evaluation and clustering of the available texts. Compared with purely lexical approaches, this procedure enables contextualized topic and emotion analyses across platforms and time periods, including frames (problem, cause, evaluation, recommendation) and narrative building blocks (roles, conflicts, proposed solutions).