Why is it that big data applications in the film industry are mostly gimmicks?

In recent years, there have been a lot of big data in the film and television cases, the word Netflix is estimated to let everyone ears to listen to the cocoon, another British script semantic analysis of the company Epagogix also become a small hot word. Mysterious Big data and mysterious film industry, after the combination, it is plated with the metaphysical color, it seems you today in social media sneeze, is worried about whether you have a cold, I will be able to make a horror film on avian flu theme tomorrow, and let you have to see the impulse.

In fact, people in the video industry know that the so-called "card House" success, rather than as a technical development case to analyze, as a PR case to learn. CMO Kelly Bennett, who was originally working in Warner, and Cco Jonathan Friedland, from Disney, made Netflix more like a TV station than an internet company, which actually made Netflix's imagination bigger, Because the internet already has a YouTube, but not a HBO or Showtime. As for the combination of Fincher, Spacey, and the 1990 solitaire, which can produce wonderful chemical reactions, this seductive idea is a means of Hollywood brokerage, rather than a language system of large data. Behind the card house, Fincher and Spacey belong to a brokerage firm whose main driver is the CAA, one of Hollywood's shadow aides, and the producer MRC, which Netflix can join just because they have more money than cable. The role of large data in the entire play will not be more valuable than the intern who recommended the British drama "The Card House" to Modi Wiczyk, the MRC co-chairman, in 2008. The services offered by Epagogix are not new, and Hollywood has thousands of script doctors, cheap and easy to use. After the 1980s, Hollywood's audience structure and viewing preferences gradually stabilized, the audience preferences is not a problem, the current in North America commonly used focus interview form, can get more concrete and practical feedback. From a certain point of view, Hollywood is not lack of good ideas and good script, the process of translating words into images is more complicated than the public information in some entertainment reports. If based solely on the angle of the script and the cast, we can predict the future, just like we are by analyzing the body of hens, and then to judge the popularity of a plate of tomatoes scrambled eggs, they are bound to each other, but this correlation is not enough to prove that a bubble surface for a living otaku, Why would immediately fall in love for his scrambled eggs mm. Why is big data hot in recent years? Humans enter the age of large data, similar to biology ushered in the microscope, astronomy found the telescope, because the network transmission and computer storage computing power, gave us a magnifying glass of information, from now on the observation of the phenomenon into a new field. In fact, since ancient times, there are multidimensional data mining behavior, the process of making the calendar may be a good example, now and then occasionally there will be about the forest Marshal Zhuge Jun Division of the legend, since the birth of computer technology, the use of data and processing has been in the synchronous development, whether distributed processing or parallel processing, Not a day jumping to today's technical heights, many of our scientific discoveries areIt was only in the last 30 years that it was done. But why is "big data" suddenly hot in these years? The reason is actually very simple, global smartphone popularity. With the improvement of the information processing capability of mobile terminals, the interface with users is not only more viscous, but also realize full time interaction, at this point, everyone's mobile terminal actually becomes a data recorder. It is more personal than the information the PC can get, exposing not only the details of the person's life, position, but also his consumption habits, the first time humans have so many data producers. Each metadata can be directly linked to a specific amount of expenditure, each number can be quantified by the currency, the commercial value of large data and each enterprise's revenue can almost directly linked. So, around "big data", stories quickly become the mainstream of the moment. But behind the hype of the media, you'll find that domestic research into complex systems is still at a stage where concepts are larger than applications, and most industries have no way of mastering data relationships in linear, closed systems, let alone translating large data into valuable information. And in the film and television industry, the industrialization system is at the beginning of the stage, many practitioners even financial statements such basic data can not understand, to understand the value of large data is somewhat unimaginable. Large data in the domestic film and television industry significance of large data technology as a tool, its application direction, nothing more than three aspects, one is too and once understanding, the second is to the future of cognition, and the third is to judge the present and real-time processing, film and television industry large data technology application if want to have great development, Then there are some problems to be solved in these three aspects. The past and once understanding of what has happened is to judge, it will involve data collection, this part will lead to controversy, the central issue is: In the end how big is called, GB or TB,PB or EB? If I want to know that "to us will eventually die of youth" this almost can be classified as a film, why in the first half of 2013, after "West Tour", I should only use social media dissemination efficiency of data mining, Or is it going back to the youth of the original novel and the memory aroused by the movie propaganda? The scope of data mining in the context of social media, then through a film promotion process, we can easily through the data to make a detailed to the minute reference, and the audience will be attracted by what kind of promotional content, but it is still just describing the appearance. The ultimate cause and effect will not be found if the observation stays in sight. We have to demonstrate the mechanism that drives the phenomenon, so what volume do we use to store and analyze the memory of our viewers to find the relationship between personal experience and collective resonance? In this respect, if only using social media data for correlation analysis, in fact, and our daily perceptual deduction is not much different, or even less than perceptual deduction flexible, it is easy because the data is not fully committed to the "Black Swan" type of error (Before the discovery of Australia, the West considered only the White swan. It has to be traced back to the larger external data of the Genesis phase, such as the consumption preferences and social experiences of the major audiences over 10 years, and the statistical analysis of their psychological activities prior to their viewing. Would it be too complicated? But from the point of view of data mining, it is only in this direction that efforts can provide substantive value. Or, we can be simple and rough some, such as Matt's director Chen Shizhi said, "to youth" the success of the most important factor is because Zhao Wei's dedication and devotion, as well as her personal accumulation in the industry. You can think about the above three angles, which will be more accessible to the core of the whole event.   In the future and the next large data technology can let human understanding of the phenomenon of more in-depth research, but when the domestic film and television project prospects for prediction, the first thing to face is that we are still in a period of continuous fluctuations in the audience group. In the North American market, the contribution of 50% of the box office audience accounted for about 10% of the population, that is, about 30 million, this group is basically a fairly stable structure. In the late 70, when the average production budget for North American films began to climb to more than $10 million, and when the cost of public spending reached more than 5 million, the monitoring of the audience from the stage of the investigation gradually into a normal monitoring. At a time when computers were only a myth, "Hollywood" created the earliest large-scale audience research model in the form of artificial + letters, which, through decades of accumulation, had made the connection between a film and the audience very transparent. But even so stringent market monitoring, in recent years, because of the impact of mobile interconnection, the audience to the cinema to watch the random behavior of the movie gradually increased, resulting in the traditional audience research model frequently appeared some problems. In the domestic film market, almost two-thirds of the screen is in the last three years, the 2010, we have a modern screen but only 6223, and now this number is almost 17000. It is conceivable that the cinema is now ushered in the audience, is basically nearly three years to gradually cultivate the interest of cinema viewing, this behavior can not be called a habit. Therefore, the current volatility of the Chinese film market is difficult to complete monitoring through existing technical means, because of the existence of other unknown variables, which lead to the opposite result, which is called "Variable omission deviation" in statistical regression analysis, the scope of large data technology can not help us solve this problem. We also need time to constantly revise our understanding of the market data, and the audience will need time to constantly develop the habit of watching the movie. In the first half of 2013, almost all practitioners were overly optimistic about action-oriented films, and in the second half, all the practitioners, including myself, were overly bullish about the movies that dominated the love element. This change in consumer behavior has been fed back to data that can be crawled from large data monitoring in some companies, but we don't know what the impact isQuantitative。 That means we can see the trend, but it's hard to determine the outcome. So, in today's Chinese film market, we might as well be more realistic about the direction of application of large data technology, from the prediction of the future to the mistakes that we can avoid. To judge the present and deal with it in real time now the understanding of large data tends to be entangled in the first word "big", while ignoring its other important feature "fine", in fact, the latter is the most important, because it will create large data real practical value. Data mining based on social media, in fact, we can make the audience classification from the simple age, gender, occupation and other dimensions, implementation to the region, activity space and character characteristics, and so more rich details, on this basis, we have to do is how to provide the audience with personalized impact, Instead of a film-oriented common publicity. For example, when a male audience sees a movie poster on a medium, it may be a long leg and a small butt, but a female audience at the same time to contact the media, what can be seen is an uncle selling Meng. When the trailer for the position promotion was put into a Hollywood blockbuster, it may be used mainly to render emotion or highlight funny, but the same more than a minute, the video site on the online trailer, is divided into several versions, to correspond to each click behind the user's profile. In this way, the audience will join the production process, through the rapid processing of audience preferences, and ultimately create more appropriate to disseminate information. At present, the data research company participates in the movie promotion the process, does is still only a statistic work, the decision-making is in the film side or is the public relations company, actually may the decision-making mechanism and the data also refine, becomes the real-time interaction, reduces the time the loss, enhances the movie promotion efficiency. We used to be in the promotion of the film, often for how to take care of the interests of the majority of the audience headaches, then a new way of thinking, with the existing audience data for the group subdivision, to different audiences to provide the same information, Haina rivers than the radiant may be more in line with the essence of the current social marketing. However, all these are only idealized vision, the reality is that China's film industry is still in a very primitive state. In terms of the cost of film investment alone, the film currently being released has an average investment of about 30 million yuan, less than $5 million trillion, which is equivalent to the level of North America in the early 70, regardless of inflation and audience income. In the face of such a market environment, many issues are actually more empty, because the head to make decisions, although there is a great risk, but after all, the cost is very low.