Big data helps movies pre­dict hits

Tech­nol­ogy ser­vice provider Paip­i­anbao gained 4,500 cinema sub­scribers through­out China

China Daily (Canada) - - BUSINESS - By MENGJING and ED ZHANG in Shaox­ing, Zhe­jiang province

Zhang Xiaom­ing is not a big fan of movies but he can tell which films will be a big hit in China and which will not by us­ing an ad­vanced tech­nol­ogy named big data.

“With enough data, many things can be pre­dicted, even the stock mar­ket in­dex,” said Zhang, who had around 20 years of ex­pe­ri­ence in data anal­y­sis in the United States be­fore head­ing the big data re­search cen­ter at Zhi­jiang Col­lege of Zhe­jiang Uni­ver­sity of Tech­nol­ogy in 2011.

In Shaox­ing, a city in eastern China’s Zhe­jiang province where the col­lege is lo­cated, Zhang and his 18-mem­ber team work on big data solutions that can­not only help pro­duc­ers know which com­bi­na­tion of cast can bring more box of­fice sales be­fore shoot­ing, but can also help cine­mas make more money by smartly ar­rang­ing the screen­ing of movies.

“Around 80 per­cent of the movies pro­duced in China can­not get no­ticed by cinema man­agers due to their poor qual­ity. Less than 10 per­cent of movies can ac­tu­ally make a profit,” said Zhang, ad­ding that it is im­por­tant to use tech­nol­ogy to help pro­duc­ers and cine­mas in de­ci­sion­mak­ing so that they can max­i­mize their prof­its.

Zhang, who re­ceived his PhD and master’s de­gree from theMas­sachusetts In­sti­tute of Tech­nol­ogy, said by us­ing cine­mas’ pre­vi­ous tick­et­ing data, data from the me­dia and the public’s re­views of movies com­bined with the al­go­rithm they de­vel­oped, his team can pre­dict the box of­fice and in­crease the num­ber of movie­go­ers to cine­mas by 10 per­cent to 15 per­cent.

He used Hong Kong film di­rec­tor Stephen Chow’s lat­est movie Mermaid as an ex­am­ple. “The movie was about to re­lease with two other movies dur­ing the Spring Fes­ti­val. All of the three movies seemed equally good by any stan­dards,” Zhang said.

“But one lo­cal cinema lis­tened to our ad­vice to use its big­gest screen­ing room to showMer­maid, and it earned hun­dreds of thou­sands of yuan within a week,” he said.

It turned out that the Stephen Chow’s comic fan­tasy not only beat lo­cally made Mon­sterHunt andHol­ly­wood ac­tion movie Fu­ri­ous 7 dur­ing the same pe­riod, but be­came the first Chi­nese movie to reach the new land­mark of 3 bil­lion yuan in gross re­ceipts in China.

Launched about 18 months ago, Zhang’s big data ser­vice named Paip­i­anbao has gained 4,500 cinema sub­scribers. China has a to­tal of 7,000 cine­mas.

The ma­jor­ity of cine­mas hire ex­pe­ri­enced man­agers to draw up their screen­ing cal­en­dar, de­cid­ing which movies will be shown in which room at which time. With Paip­i­anbo, the cal­en­dar can be au­to­mat­i­cally compiled within one minute.

“The ser­vice re­places the role of cinema man­agers to some ex­tent. The most dif­fi­cult as­pect of fur­ther ex­pand­ing our busi­ness is how to con­vince more cine­mas to adopt our tech­nol­ogy,” he said.

He said that his ser­vice has the most ac­cu­rate pro­jec­tion on box of­fice so far but there are other play­ers which want to repli­cate his suc­cess. “I don’t think their pro­jec­tion can be as ac­cu­rate as ours. But box of­fice pre­dic­tion is not a 100-me­ter race, even if you run faster than your com­pe­ti­tion by sec­onds, it doesn’t make a big dif­fer­ence,” he said.

Con­tact the writ­ers through mengjing@ chi­nadaily.com.cn

Newspapers in English

Newspapers from China

© PressReader. All rights reserved.