The term "automated journalism" refers to the use of NGL algorithms (Natural Language Generation) for the rapid and efficient production of journalistic pieces. Large editorial organizations, such as Associated Press, The Washington Post, and Los Angeles Times, among others, use them systematically to produce journalistic pieces from a set of structured data, which the algorithm can obtain either from the Internet or from various templates filled out by journalists, or from large databases. Spanish newspapers such as El Confidencial have tried some for sports contexts (one of the fields most worked on by the algorithm for the ease of obtaining information and structuring these pieces).

This is an innovation called to reinvent journalism and information processing, especially in a context where immediacy and the need to always be publishing is evident.

This project is born as an attempt to think about this phenomenon, to analyze it, to study the perception of professionals. It is also a gateway to rethink the journalistic function, from teaching and from praxis, deontology, and the principles that have configured the competence prototype of the journalist for years. In this sense, based on the results obtained and in connivance with different organizations that are working in this line, it could be explored how the results can modify or improve these software, templates, or information processing mechanisms

This emerging field has publications analyzing various aspects, from the level of trust of these products to the perceptions of journalists. Improvements in structuring processes have also been proposed to contribute to obtaining products with a more coherent and "human" narrative structure. In the Spanish context, there has been very little academic production on the subject. Torrijos and Bran (2019) analyzed the functioning of an algorithm produced by El Confidencial to make sports chronicles based on a content analysis. Túnez-López, Toural-Bran, and Valdiviezo-Abad (2019) conducted a bibliographic study on automated journalism, although they did not obtain experimental data. The limited experimentation is extremely recent and concise: Noain-Sánchez (2022) conducted 15 in-depth interviews to analyze the perception of journalists and experts. From the UPF, Lluís Codina and Mari Vállez (2018) published a bibliographic study on computational journalism where these algorithmic innovations were mentioned, although they did not focus on the study of the algorithms or their perception, but on a broad sense of computational innovations applied in newsrooms (from image processing to other digital tools).

Recent studies reflect on cases of RTVE, which wants to use algorithms to cover the 2023 municipal elections (Aramburú, López, López, 2023) or studies on the development of technology in the case of journalism in Spain (Sánchez-García, Merayo-Álvarez, Calvo-Barbero, and Diez-Gracia, 2023).

The main objective of the project is to: 

• To analyze, using a mixed methodology, the perception of a representative sample of specialized journalists on automated journalism.