Predicting Flu Trends using Twitter Data使用推特数据预测流感趋势.pptVIP

Predicting Flu Trends using Twitter Data  Harshavardhan Achrekar [1] Avinash Gandhe [2] Ross Lazarus [3] Ssu-Hsin Yu[2] Benyuan Liu [1]  Workshop on Cyber-Physical Networking Systems in conjunction with INFOCOM 2011CPNS 2011, Shanghai, China SNEFT – Social Network Enabled Flu Trends [1] Computer Science Department, University of Massachusetts Lowell[2] Scientific Systems Company Inc , Woburn, MA[3] Department of Population Medicine - Harvard Medical School Background Related Work Our Approach SNEFT System Architecture Twitter Dataset Description Twitter Dataset Analysis Detection and Prediction Conclusion Outline Seasonal flu Influenza (flu) is contagious respiratory illness caused by influenza viruses. Seasonal - wave occurrence pattern. 5 to 20 % of population gets flu ≈ 200,000 people are hospitalized from flu related complications. 36,000 people die from flu every year in USA. worldwide death toll is 250,000 to 500,000. Epidemiologists use early detection of disease outbreak to reduce no. of people affected. Related Work :- Google Flu Trends  Certain Web Search terms are good Indicators of flu activity. Google Trend uses Aggregated search data on flu indicators. Estimate current flu activity around the world in real time. Accuracy of data {not every person who searches for “Flu” is sick} From example :- Google Flu Trend detects increased flu activity two weeks before CDC. CDC stands for Center for Disease Control Link:- /flutrends Our Approach OSN emerged as popular platform for people to make connections,share information and interact. OSN represent a previously untapped data source for detecting onset of an epidemic and predicting its spread. {“i am down with flu”, “got flu.”} msg exchange between users provide early ,robust predictions. Twitter/Facebook mobile users tweet/posts updates with their geo-location updates. helps in carrying out refined analysis. User demographics like age, gender, location, affiliated networks.,etc can be i