Adaptive nowcasting of influenza outbreaks using  Google searches

Abstract

Seasonal influenza outbreaks and pandemics of new strains of the influenza virus affect humans around the globe. However, traditional systems for measuring the spread of flu infections deliver results with one or two weeks delay. Recent research suggests that data on queries made to the search engine Google can be used to address this problem, providing real-time estimates of levels of influenza-like illness in a population. Others have however argued that equally good estimates of current flu levels can be forecast using historic flu measurements. Here, we build dynamic ‘nowcasting’ models; in other words, forecasting models that estimate current levels of influenza, before the release of official data one week later. We find that when using Google Flu Trends data in combination with historic flu levels, the mean absolute error (MAE) of in-sample ‘nowcasts’ can be significantly reduced by 14.4%, compared with a baseline model that uses historic data on flu levels only. We further demonstrate that the MAE of out-of-sample nowcasts can also be significantly reduced by between 16.0% and 52.7%, depending on the length of the sliding training interval. We conclude that, using adaptive models, Google Flu Trends data can indeed be used to improve real-time influenza monitoring, even when official reports of flu infections are available with only one week's delay.

Analysis of the communities of an urban mobile phone network

“…The increasing pervasiveness of always-on technology creates a vast amount of information that closely reflects human activity. This provides insight into the behaviour of people across the levels of their environment, from the individual scale, through groups and communities, to the global sphere, enabling the creation of models with predictive power [3–5]. Data recorded from mobile phones are a target of choice for research of this kind, offering a high granularity and being effectively ubiquitous in our society [6–12].…”

A Synthesis of Spatial Models for Multivariate Count Responses

“…In terms of emerging opportunities, a potentially transformative one is seen in extending advanced spatial models in settings that use geo-referenced, real-time input data to make forecasts about current or near-future values (i.e., nowcasting [e.g., Lampos et al 2015, Preis andMoat 2014]). Recent years have seen a rapid growth of real-time data with location attributes, from Google's influenza reports (which exploit Internet users' search queries), through pedestrian or cyclist route and volume data collected from smart-phone applications (Smith 2015), to vehicle and driver information streamed from connected and instrumented vehicles.…”

Early Signs of Financial Market Moves Reflected by Google Searches

“…Studies in this area have demonstrated promising links between aggregate online behaviour and collective real world behaviour across a range of data sources such as the search engine Google [14][15][16][17][18][19][20][21], the search engine Yahoo! [22,23], the online encyclopedia Wikipedia [24][25][26][27][28], the microblogging platform Twitter [29][30][31][32][33][34] and the photosharing website Flickr [35,36].…”