Data to text for generating information of weather and air quality in the R programming language

Data obtained from weather monitoring stations and air quality acquired from several meteorological websites are numerical data. The research is aimed at developing a weather-generating Data to Text (D2T) that enables people to understand weather and air quality information. On developing D2T, there are four basic elements: signal analysis, data interpretation, document planning, as well as microplanning and realization. This research contributes to extending the signal analysis by adding new features: exponential smoothing for prediction time series data, a linear model with gradient descent for handling missing values and data summarization with statistical tools. In this study, several packages available in the R programming language were utilized. To validate the results, four measuring tools consisting of readability, computation time, relevance and truthfulness and comprehensibility and importance were used. After evaluating the results, it was stated that on the readability aspect the developed system can be understood by students at the age of 13-15 yearscomputational time is low (round 2.64 s) and sufficient scores for relevance and truthfulness and comprehensibility and importance were obtained. Therefore, the proposed system can be used to generate text for describing weather news and air qualit