Sun, Feb 04, 2018 - Page 6 News List

Taiwanese leaders still strangers to big data

By Weber Lai 賴祥蔚

As an amateur politician, Ko Wen-je (柯文哲) won the Taipei mayoral election in 2014. Public opinion attributed a large part of the win to “big data.”

Overnight, big data took the national and local governments by storm, and it suddenly became a political hot topic. However, an examination of how governmental bodies use big data surprisingly shows that the government does not understand it.

Big data opens a wide range of possibilities. In terms of public opinion research, big data can be used to engage in social listening — monitoring social media channels or digital conversations for mention of a certain issue — through self-developed software or by commissioning professional organizations.

Institutions specializing in social listening have developed user-friendly interfaces that are as easy to use as any common software.

If governments use social listening to gain an understanding of public opinion, it should be done by a department set up for the purpose. In that way, a one-time investment would allow the whole agency or local government to gain a thorough understanding.

Disappointingly, some in government still monitor public opinion on the Internet manually and some individual agencies have commissioned different specialized organizations to handle big data for them. This is more expensive and means that results are not integrated.

It is as easy to collect online opinions as it is to conduct an opinion poll. However, casually conducted polls are never good representations of public opinion, and the same is true for big data analysis and social listening, which are even more instantaneous.

The quality of the programming, the completeness of the database and the accuracy of the semantic analysis are all essential factors when it comes to listening to public opinion on the Internet. In addition, filtering out the fake opinions overflowing the Internet is another hot topic in social listening.

Many instances of fake public opinion were uncovered last year: In a report submitted to the US Congress, Facebook said organizations had been promoting certain values by using fake personal accounts and that they were all Russia-backed; the BBC last year reported that public opinion in Brazil during the previous general election was influenced by fake social media accounts.

Some Internet users were paid to simultaneously run more than 20 fake personal accounts per person.

Taiwan is still in the early phases of detecting fake social media accounts and forged public opinion, according to Willie Li-wei Yang (楊立偉), who has been studying big data for years and is now managing director of eLand Information Co.

Yang has said that examining inconsistencies in user status and posting behavior can help filter out 60 to 70 percent of fake accounts.

When looking at current practice in social listening, it is evident that the government is still incapable of putting big data analysis to good use, let alone conducting further research and development.

There will be local elections in November, and a solid understanding and proper application of big data technologies would be helpful to candidates who want an up-do-date analysis of public opinion trends — in addition to using their own perceptiveness and political judgment, and relying on poll results.

If leaders in the government continue to be strangers to big data instead of making an effort to learn how to use it more intelligently, they will be swept away by the tides of time.

Comments will be moderated. Keep comments relevant to the article. Remarks containing abusive and obscene language, personal attacks of any kind or promotion will be removed and the user banned. Final decision will be at the discretion of the Taipei Times.

TOP top