Search engine to use online queries to predict health trends before official health bodies

Google already has a window into our souls through our internet searches and it now has insight into our ailing bodies too.

The internet giant is using its vast database of individual search terms to predict the emergence of flu up to two weeks before government epidemiologists.

Google Flu Trends uses the tendency of people to seek online help for their health problems.

By tracking searches for terms such as 'cough', 'fever' and 'aches and pains' it claims to be able to accurately estimate where flu is circulating.

Google tested the idea in nine regions of the US and found it could accurately predict flu outbreaks between seven and 14 days earlier than the federal centres for disease control and prevention.

Google hopes the idea could also be used to help track other diseases. Flu Trends is limited to the US.

Jeremy Ginsberg and Matt Mohebb, two software engineers involved in the project, said: "Patterns in Google search queries can be very informative."

In a blogpost on the project they wrote: "It turns out that traditional flu surveillance systems take one-two weeks to collect and release surveillance data but Google search queries can be automatically counted very quickly.

"By making our estimates available each day, Flu Trends may provide an early-warning system for outbreaks of influenza."

They explained that private information health would be kept confidential. " Flu Trends can never be used to identify individual users because we rely on anonymised, aggregated counts of how often certain search queries occur each week."

A paper on the project has been accepted by the respected journal Nature.

"This seems like a clever way of using data that is created unintentionally by the users of Google to see patterns in the world that would otherwise be invisible," Thomas Malone, a professor at the MIT Sloan school of management, told the New York Times.

在官方的健康組織還沒有發布健康趨勢之前,搜索引擎就能利用網上的搜索查詢內容來預測它。

在我們用Google進行網上搜索的時候, 她已經打開我們心靈的窗戶，現在她又要洞察我們的病體。這個互聯網的巨人正在利用她收集到的無數個人搜索詞匯數據趕在政府流行病學家之前兩個星期來預測流感的出現。

Google流感趨勢項目就是利用了人們在網上對他們的健康問題尋求幫助的趨勢來預測流感。通過追蹤像“咳嗽”，“發燒”和“疼痛”這樣的詞匯，它表明已經能夠準確地判斷流感在哪里擴散。Google在美國的九個地區就這一觀點做了測試并且發現她比聯邦疾病控制和預防中心提前了7到14天準確預測了流感的爆發。Google希望這一發現同樣能幫助預測別的流行病。流感預測目前也只是限于在美國。

參與這一項目的兩名軟件工程師Jeremy Ginsberg 和Matt Mohebb 說, “ Google搜索顯示的數據分布模式非常有價值。” 他們還在關于這一項目的日志中寫到: 結果是傳統的流感監測系統要用一到兩個星期來收集和發布監測數據而Google搜索查詢統計卻是在很短的時間內自動完成的。“通過我們每天的評估，流感趨勢項目可以為流感的爆發提供一個早期預警系統。”

他們還解釋說，會為私人的健康信息保密。“流感趨勢項目絕不用來透漏個人使用者的身份，因為我們依靠匿名的匯總的計數來知道一定的查詢內容在每個星期發生的頻率。”

著名期刊《自然》已經準備刊登關于這一項目的論文。麻省理工學院斯隆商學院的教授Thomas Malone在接受《紐約時報》采訪時說,“使用Google 用戶無意創造的數據來看世界上原本看不見的模式,這看起來是一個聰明的做法。”