The new system allows regulators to capture diverse viewing needs. Photo: FILE

---

 

The National Radio and Television Administration (NRTA) launched a comprehensive evaluation system for television stations recently. The new system, based on independent technology, can collect live data on a cable television audience of 22 million and nearly 20 million IPTV users across the country, which will help close loopholes in the traditional model for measuring viewership figures. 

Du Baichuan, deputy director of the Science and Technology Commission at NRTA, said that China used to adopt two practices common in the world, namely daily record cards and audience rating measuring equipment. In some cases, sampled users would fill in the cards every day to record their chosen programs, after which, investigators would collect the cards door to door for further analysis. In other cases, sampled users were equipped with a remote-control instrument similar to a television set-top box. After turning on TV, they would press in the code representing their personal information and watch a television channel for a while, and then the data could be recorded and transmitted to investigation companies.

These two models are both small-scale surveys that greatly weaken the credibility of audience ratings. For example, if 5 million users in a city are sampled, then the programs they have not watched will be counted at zero percent. In the new system, however, even a 1 percent viewership rating means that 5,000 viewers watched the program for a minute.

In addition, it is difficult for the traditional surveys to adapt to new situations due to technological development and diversified methods for watching programs. Du said that viewers not only watch programs on television, but also on computers, tablets and mobile phones. Audiences have more choices in terms of platforms, including cable television, satellite television, IPTV and internet television. They can watch live broadcasts, and they can also appreciate the program through playback functions.

Considering this trend, the new system has been based on big data and cloud computing technology. It can collect real-time statistics from the whole sampled audience, and its analysis can be as precise as an individual household. In this way, the result reflects the viewership of popular programs and prime time while still accurately presenting the characteristics of the programs that receive less attention, thus capturing diverse viewing needs.

Yu Ying, dean of the Radio and Television Planning Institute at NRTA, said that the new system will reach hundreds of millions of samples in the future based on live broadcasting, playback review and on-demand broadcasting. The massive data will track users’ diverse need for content and methods of viewing radio and television programs.

Authenticity is the bedrock of audience rating data and analysis. However, the falsification of ratings has long been a phenomenon that the TV industry has detested but found hard to eradicate.

According to international practice, the names and addresses of sampled viewers should be kept strictly confidential in order to avoid television stations, program producers and advertisers’ bribing the audience and doctoring data. Wang Pengju, secretary-general of the China TV Production Industry Association, held that although the previous survey method was a scientific method based on statistics, it had enormous technical loopholes for ratings fraud due to its small samples and uneven distribution. Audience ratings thus became vulnerable to interference and easy to cheat.

“Ratings fraud has turned into a production chain. Advertising is the main source of revenue for TV stations, and audience ratings are a major factor for advertisers to consider. TV stations and producers are spending millions to buy off sampled users’ information. It is difficult to punish such fraudulent acts due to a lack of explicit legal terms,” said You Xiaogang, president of the China Television Drama Production Industry Association.

Compared with the previous method, the new system based on big data technology is obviously more convincing. In the system, data collection, cleaning, analysis and presentation can be completed consistently through automation and closed processing. Its protections against manipulation have greatly increased the cost of data fraud. The system depends on massive big data statistics, so the impact of individual sample fraud on statistical results is almost negligible.
Yu pointed out that the new system is guided and supervised by the NRTA, which also ensures data legitimacy and fairness to a greater extent. At the same time, the NRTA is striving to optimize relevant laws and regulations.

Du argued that artificial intelligence applications based on the new system will be realized in the near future. “In the past, people thought that data and big data were the same thing, but this is not the case. When we talk about big data, we are talking about understanding and analyzing data by various factors related to each other,” Du said, noting that “it is on this basis that the new system functions. Its ability to apply new technology is much stronger than the traditional statistical method. Now we are watching TV, and TV may watch us in the future. Eyeball tracking technology may be part of the system and the sound control data may be better utilized. This allows us to picture future life and greater possibilities. Also, this is what matters to the new system most.”

 

This article was translated from People’s Daily.

​(edited by MA YUHONG)