Linguistic Descriptions for Automatic Generation of Textual Short-Term Weather Forecasts on Real Prediction Data

Abstract

We present in this paper an application which automatically generates textual short-term weather forecasts for every municipality in Galicia (NW Spain), using the real data provided by the Galician Meteorology Agency (MeteoGalicia). This solution combines in an innovative way computing with perceptions techniques and strategies for linguistic description of data together with a natural language generation (NLG) system. The application, named GALiWeather, extracts relevant information from weather forecast input data and encodes it into intermediate descriptions using linguistic variables and temporal references. These descriptions are later translated into natural language texts by the natural language generation system. The obtained forecast results have been thoroughly validated by an expert meteorologist from MeteoGalicia using a quality assessment methodology which covers two key dimensions of a text: the accuracy of its content and the correctness of its form. Following this validation GALiWeather will be released as a real service offering custom forecasts for a wide public.

Adapting SimpleNLG to Spanish

“…Secondly, we tested SimpleNLG-ES in the real data-to-text service GALiWeather (Ramos-Soto et al, 2015), which is a template-based NLG system that was deployed in May 2015 as a public service for the Official Meteorology Agency (MeteoGalicia, 2000) of Galicia (NW Spain). GALiWeather produces automatically daily operational weather forecasts for each of the 314 municipalities in Galicia.…”

On the role of linguistic descriptions of data in the building of natural language generation systems

“…Natural language generation (NLG) is described by John Bateman in [13] as the branch of natural language processing which deals with the problem of how texts in human natural language can be automatically created by a machine. This may be seen as the inverse of the problems addressed by natural language understanding but, actually, the NLG field emerges from a very different set of motives and objectives, both theoretical and practical.…”

“…In this sense, we believe the best way to dive into this research field is through the book by Reiter and Dale [4], which addresses the whole NLG problem and describes the design and development of a full system in detail through a thorough example. The review by Bateman [13] also provides both introductory and deep reflections about this field, including a thorough list of NLG approaches which covers from the beginnings of the field to the early 2000s.…”

“…Due to its short career and its formal background, many approaches in this field are on the theoretical side, although in some cases practical examples and real life based problems are given. More recently, the use of hybrid approaches which employ LDD techniques together with NLG systems to provide solutions to real life problems has emerged [12].…”

Proceedings of the 10th International Conference on Natural Language Generation

“…Topically limited systems are already being used in different areas like weather reports (Ramos-Soto et al, 2013) or financial analysis (Nesterenko, 2016). Our topically unlimited software has an abstraction layer for text planning, structure, semantics, and content that can be fitted to any kind of subject.…”