Perspectives and challenges of artificial intelligence techniques in the field of social sciences and communication

Abstract

In the last ten years, artificial intelligence (AI) methods and techniques have witnessed great advances and they already make part of the usual landscape from where new or old problems are tackled in different areas of human knowledge. Three particular aspects are behind this leap forward: A generalized availability and variety of data; a deeper understanding of the mathematics governing the underlying algorithmics; and hardware capabilities allowing wide and deep experimental pipelines over data. The main challenge in each problem and context of application now lies on understanding how can these technologies can be used, their reach and limitations so that the can be aligned with the aims of each specific problem at hand. Social and communication sciences are no exception, but show particularities that define which AI technologies and methods are most appropriate (i.e. natural language processing). This works presents an introduction to the methodology under which AI models are built, to potentially useful ai services in the field and, finally, some examples of applications illustrating practical and technical considerations in this respect.