Entertainment brain really can have "smart"? The answer is the life cycle of artificial intelligence technology! From the content intelligence to increase marketing, entertainment panoramic Secret Ali intelligent algorithm, in this one!
Click here to get free e-books
Ali is a strength of the company believes the data, entertainment and related industries involves a wide range, from online to offline, from theatrical comprehensive diffuse to live entertainment and literary fiction, its composition, form, show the complexity of interwoven distributed in with tremendous challenges of business data.
The past three years, Ali Civic Mount Cool Lab has always been to help business development and growth as the core driver, form the technical direction of the following four:
Content understanding is the cornerstone of the contents of different forms, IP, novels, plays, video, music and other entertainment-related algorithm technology to build knowledge map of the field brought a lot of difficulties, in which computer vision, natural language processing, maps & reasoning, FIG neural network, multi-modal content analysis technology is widely applied to the content deconstruction. Video, for example, the length of motion picture video is difficult to sum up with some of the low-level abstraction to express its label content, based on multi-modal analysis techniques will run into a wall on the type of content, content integration and machine learning expert system semi-automatic micro-label system as a viable way out. And a short video online fast feedback loop different, even the shortest production cycle variety show also takes three months or more, still faces uncertainties with monitoring the audit period, which led to theatrical comprehensive content production high degree of uncertainty, how to evaluate selected content based on complex analysis of historical data leads and the success or failure of the law is one of the core challenges faced by the various integrated video platform, and the Big Dipper Ali entertainment system is used to solve this problem.
Search and recommendation as an important means of information explosion two solutions are widely used in various APP, rather theatrical comprehensive content the user wants to lead the complexity of a precise description of the content is very difficult to retrieve only to the user through the program name, actor name also It caused a great deal of distress. In entertainment content distribution system in the search mode, the recommended mode of fusion as a new user requirements, how to strengthen the class through a more accurate understanding of user intent learning process to help them find favorite content as soon as possible, become entertainment system pushed down search the primary task of a phase.
Entertainment industry as an important development of the Internet industry, artificial intelligence, space technology applications in the field of the majority, and we are just a small step, looking forward to engineers to create greater miracle, accelerate the industrialization era of digital entertainment industry 's arrival
