Online Social Networks Flu Trend Tracker: A Novel Sensory Approach to Predict Flu Trends

Liu, Benyuan; Achrekar, Harshavardhan

doi:10.1007/978-3-642-38256-7_24

Cited by 19 publications

(19 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Timeliness of web query data is not an issue, and nowadays can simply be downloaded in real-time (from Google Flu [13] and Dengue [14] Trends). The same hypothesis has been applied to Twitter postings, with impressive results [15,16,17,18,19,20].…”

Section: Introductionmentioning

confidence: 99%

“…We will evaluate the correlation between time-series of disease-related media reports and ILINet data; we expect this correlation to be smaller than the 90% Pearson correlation coefficients observed for Twitter [17,18]. We will explore time-series methods, based on autoregressive moving average models with exogenous inputs (ARMAX models), to nowcast disease activity i.e., ILINet data, with HM time-series acting as a guide ("exogenous input").…”

Section: Introductionmentioning

confidence: 99%

“…While ARX models (autoregressive models with exogenous inputs) are quite common in econometrics [11], their use in nowcasting ILI activity using social media data has only been performed with Twitter as the exogenous datastream [28,18]. Again, while the technical challenge primarily lay in detecting influenza-related tweets i.e., in text mining, an ARX model was considered.…”

mentioning

confidence: 99%

“…Again, while the technical challenge primarily lay in detecting influenza-related tweets i.e., in text mining, an ARX model was considered. When applied to CDC ILINet data from the 2010-2011 season [18], it was found that a model that employed 2 weeks of lag in the dependent variable and none in the exogenous one (the fraction of influenza-related tweets, collated on a weekly basis) provided the most predictive skill. Interestingly, the dependent variable was a logit-transformed ILI fraction, whereas an earlier work by the same authors [28], applied to the tail-end of the swine-flu pandemic in 2009-2010 used ILI fractions i.e., without logit-transforms, and found that autoregression degraded the model's predictive skill.…”

mentioning

confidence: 99%

“…This results in multiple "calibrated" models, the most predictive of which needs to be selected for further use. This may be detected via cross-validation (out-of-sample tests) as performed in [18] or, since the models are fully nested, one may compute various information criterion e.g., Akaike Information Criterion (AIC), Generalized Cross Validation (GCV), Portmanteau tests, RICE criterion, Final Predictive Error (FPE) and use them for model selection [29]. Theory on ARMAX models can be found in [30,31].…”

mentioning

confidence: 99%

See 4 more Smart Citations

Nowcasting influenza outbreaks using open-source media report.

Ray¹,

Brownstein

2013

View full text Add to dashboard Cite

We construct and verify a statistical method to nowcast influenza activity from a time-series of the frequency of reports concerning influenza related topics. Such reports are published electronically by both public health organizations as well as newspapers/media sources, and thus can be harvested easily via web crawlers. Since media reports are timely, whereas reports from public health organization are delayed by at least two weeks, using timely, open-source data to compensate for the lag in "official" reports can be useful. We use morbidity data from networks of sentinel physicians (both the Center of Disease Control's ILINet and France's Sentinelles network) as the gold standard of influenza-like illness (ILI) activity. The time-series of media reports is obtained from HealthMap (http://healthmap.org). We find that the time-series of media reports shows some correlation (≈ 0.5) with ILI activity; further, this can be leveraged into an autoregressive moving average model with exogenous inputs (ARMAX model) to nowcast ILI activity. We find that the ARMAX models have more predictive skill compared to autoregressive (AR) models fitted to ILI data i.e., it is possible to exploit the information content in the open-source data. We also find that when the open-source data are non-informative, the ARMAX models reproduce the performance of AR models. The statistical models are tested on data from the 2009 swine-flu outbreak as well as the mild 2011-2012 influenza season in the U.S.A.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Nowcasting influenza outbreaks using open-source media report.

Ray¹,

Brownstein

2013

View full text Add to dashboard Cite

show abstract

Trends detection of flu based on ensemble models with emotional factors from social networks

Sun

Ren

2016

IEEJ Transactions Elec Engng

View full text Add to dashboard Cite

Influenza is an acute respiratory illness and widespread activity that occurs every year. Detection and prevention of influenza in its earliest stage would reduce the spread range of the illness. Sina microblog is a popular microblogging service in China, which can be treated as perfect reference sources for flu detection because of its real-time character. A large number of active users post about their daily life continually. In this paper, we investigate the real-time flu detection problem and propose a flu detection model with emotion factors and semantic information. First, we extract flu-related microblog posts automatically in real time by adopting support vector machine (SVM) filter and semantic features. We use association rule mining to extract strongly associated features as additional features for posts to overcome the limitation of 140 words, including sentiment information, which can help us to classify the posts without flu-related features. Then, the conditional random field model is revised and applied to detect the transition time of flu so that we can find out which place is more likely to have influenza outbreak and when it is more likely to have influenza outbreak in a city or province in China. Experimental results on detecting flu situation during certain times in some locations show the robustness and effectiveness of the proposed model, which might help health authorities in predicting flu outbreak ahead and take timely control action and response.

show abstract