Towards a Conceptual Model for Data Narratives

Data narration is the activity of producing stories supported by facts extracted from data analysis, possibly using interactive visualizations. In spite of the increasing interest in data narration in several communities (e.g. journalism, business, e-government), there is no consensual definition of data narrative, let alone a conceptual or logical model of it. In this paper, we propose a conceptual model of data narrative for exploratory data analysis. It is based on four layers that reflect the transition from raw data to the visual rendering of the data story: factual, intentional, structural and presentational. This model aims to support the entire lifecycle of building a data narrative, starting from an intentional goal: fetch and explore data, bring out highlights, derive important messages, structure the plot of the data narrative, and render it in a visual manner. Our contributions include a description of the model and its instantiation for several real examples showing that it covers data narration needs.