Can Baidu's Version Of Google Flu Trends Succeed In Predicting Flu Trends?

Jun.24.14 | About: Baidu, Inc. (BIDU)

Summary

Google Flu Trends received a lot of criticism due to flaws and inaccuracies since it was launched in 2008.

Baidu is expected to launch its own flu prediction tool in a joint venture with the local CDC.

Baidu could learn from Google’s case, use the second competitor advantage and government support to develop a successful tool.

Google Flu Trends initiative and scientists criticism

In 2008, just a few months before the 2009 H1N1 swine flu pandemic, Google (NASDAQ:GOOG) announced Google Flu Trends, a new initiative to help predict outbreaks of flu through big data analysis. Google scientists try to use health-related search phrases and users' locations in order to forecast flu activity worldwide. The idea sounds logical; people look for health-related phrases on Google when they concerned about health problems, when they want to learn how to prevent disease, and when they try to diagnose themselves. Google can aggregate flu-related search phrases, combine them with users' locations, and reveal flu presence in a specific area. Comparing this data to historical flu activity figures from US Center of Disease Control and European Influenza Surveillance Network can highlight the extent of the flu in different countries around the world. Although Google Flu Trends seems like a valid idea with great results, according to Google, just a few months after its launch, Google's algorithm failed to estimate the magnitude of the 2009 swine flu pandemic, and as a result, Google changed its algorithm to improve future estimates. Four years later, in 2013, an article published in science magazine Nature claimed that flu estimations, according to Google's algorithm during the 2012-13 flu season, were double the numbers of the US Centers for Disease Control. Earlier this year, a group of scientists at Harvard University published a paper claiming that people do not know how to diagnose flu by themselves, and typical symptoms such as fever or cough can be related to various other diseases and not necessarily to flu. The Harvard study refers to Google's attempt to determine flu trends using big data analysis as "Big Data Hubris" and suggests that, in some cases, regular data collection and analysis yields better results than web data and social network mining.

Baidu's initiative

Earlier this month, Chinese state press agency Xinhua reported that search giant Baidu (NASDAQ:BIDU) and the Chinese Center for Disease Control ("CDC") will launch a prediction tool to help forecast flu outbreaks, similar to Google Flu Trends. The tool will use big data analysis and users' locations to understand flu trends in the country that suffered from the SARS outbreak in 2003. Lai Shengjie, a researcher at the CDC's infection prevention and control department, said that they are well aware of the flaws of Google Flu Trends and will adjust the data by repeatedly comparing the data-generated trends with traditional flu surveillance results.

Baidu has a good chance to succeed

Baidu has the advantage of being the second to try to estimate flu trends through big data analysis. As the second company to offer a flu prediction tool, Baidu can use all the latest publications about flaws and inaccuracies in Google Flu Trends and include solutions for these problems in its code from the beginning, without having the need to tweak the algorithm. For example, Harvard researchers claimed that not every person looking for the words "fever" or "cough" is sick; Baidu can leverage this and combine the search phrase with recent web browsing history to receive better results. Also, Baidu and the Chinese CDC can test Google Flu Trends by themselves to find more problems that were not published and include solutions for these problems in the original code as well. Previous cases of companies using the advantage of being the second competitor to release a more popular service and gain a bigger market share than the original service can include Facebook and Myspace, Skype and VocalTec, Google and AltaVista, etc. Similar to Google, Baidu may also be guilty of big data hubris; however, in Baidu's case, locals in China do not necessarily visit doctors and prefer to wait before seeking professional advice. As the traditional flu monitoring system is based on hospital reports to the CDC, traditional data may not be accurate in these areas, and Baidu's flu prediction tool can assist in estimating the magnitude of a flu trend there. As this tool could help the Chinese government to reveal, isolate, and prevent cases of flu in these areas, Baidu will benefit from greater government support than Google Flu Trends. Since Google's launch of Flu Trends, big data analysis has evolved. More knowledge and better technology and software are available today than a few years ago when Google launched its flu prediction tool. Baidu could use the latest technology to develop a high quality tool from the beginning that will use the latest developments in the big data analysis market.

Conclusions

Baidu and the Chinese Center for Disease Control launched a joint project to use big data analysis and develop a flu trend prediction tool to help estimate flu trends in China. The announcement of the new initiative is six years after Google launched its Google Flu Trends in order to estimate flu trends around the world. Google received a lot of criticism, and a few research articles were published highlighting the flaws and inaccuracies of Google's tool. Baidu could develop a better tool if it learned from Google's case and the criticisms of it and then implement its conclusions into the new tool. In some areas of China, such tool is vital in order to spot an epidemic, as locals don't seek professional medical advice often, and the Chinese government will give its support to Baidu to succeed in this initiative. Baidu has a good chance to succeed in this initiative, and it could bring a lot of positive publicity worldwide to the search giant.

Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: Information provided in this article is for informational purposes only and should not be regarded as investment advice or a recommendation regarding any particular security or course of action. This information is the writer's personal opinion about the companies mentioned in the article. Investors should conduct their own due diligence and consult with a registered financial adviser before making any investment decision. Lior Ronen and Finro Financial Consulting and Analysis are not registered financial advisers and shall not have any liability for any damages of any kind whatsoever relating to this material. By accepting this material, you acknowledge, understand and accept the foregoing.