Forecasting with Twitter data: An application to USA TV series audience
Free (open access)
Volume 11 (2016), Issue 3
220 - 229
L. MOLTENI & J. PONCE DE LEON
Various researchers and analysts highlighted the potential of Big Data, and social networks in particular, to optimize demand forecasts in managerial decision processes in different sectors. Other authors focused the attention on the potential of Twitter data in particular to predict TV ratings. In this paper, the interactions between television audience and social networks have been analysed, especially considering Twitter data. In this experiment, about 2.5 million tweets were collected, for 14 USA TV series in a nine-week period through the use of an ad hoc crawler created for this purpose. Subsequently, tweets were classified according to their sentiment (positive, negative, neutral) using an original method based on the use of decision trees. A linear regression model was then used to analyse the data. To apply linear regression, TV series have been grouped in clusters; clustering is based on the average audience for the individual series and their coefficient of variability. The conclusions show and explain the existence of a significant relationship between audience and tweets.
Audience, Forecasting, Regression, Sentiment, TV, Twitter.