In recent years, "big data" has gradually moved from academic concept to industrial application, and has become the "window" for the Internet industry to compete for each other. However, although there are endless ideas, research and plans about big data in the industry, most companies have not actually enjoyed the change dividend brought by big data.
Nowadays, with the maturity of face recognition, artificial intelligence, cloud computing and other technologies, big data applications have also landed and begun to reshape existing business models.
Fashion can also be customized.
The favorite color of young people born after 1995 in China is "unsaturated black" with RGB value of 22/20/24. When buying clothes, the colors they most often match with black are "red, yellow and blue". The favorite clothing materials are cotton, wool and leather, and the favorite colors are printing, graffiti and letters … …
The above conclusions come from the "China Post-90s Fashion Color Report" jointly released by Tencent and e-commerce platform Vipshop. Relying on Tencent’s AI (artificial intelligence) face recognition and image processing technology, based on Tencent’s QQ space 2 trillion social media photos and Vipshop’s 100 million e-commerce users’ sales data, the two companies jointly "calculated" China’s "post-1995" dressing preferences.
It is not easy to extract the "fashion factor" from 2 trillion photos and 100 million e-commerce user data. According to Liang Zhu, vice president of Tencent Group and head of Youtu Lab, finding these popular colors needs to overcome many technical difficulties. For example, the photos uploaded by users in QQ space are mainly life photos, and the background and colors are very complicated. In order to achieve accurate color acquisition, its technical team uses many artificial intelligence algorithms, including human body detection and clothes segmentation, to separate the human body and clothes in the photos from the background, and its pixel segmentation accuracy reaches 95%.
It is understood that the two companies jointly invited designers to design fashion clothes based on the calculation results, and plan to sell them online on the Vipshop e-commerce platform this fall.
"Face recognition and image processing technology based on machine learning has always been a hot technology in the future, and its application range is very wide. At present, Tencent Youtu’s AI image recognition technology is very mature and can be applied in different scenarios. " Beam-column representation.
No longer "looking at the number and sighing"
In the past, big data applications were mostly confined to the marketing level. Its essence was to use the advantages of "big data" to analyze market factors such as user behavior by traditional computing methods, which was not a real big data application. In the face of massive and precious data, it is often impossible to really use it, and it can only be "sighing and sighing".
Nowadays, the application of big data has broken through the limitations of data capture and statistical methods in the past, and has begun to penetrate into all aspects of the industrial chain, becoming a "closed-loop ecology" that runs through big data.
In the retail field, following the concept of "new retail" put forward by Alibaba last year, at the beginning of the new year, Alibaba Group signed a comprehensive strategic cooperation agreement with Bailian Group, announcing the joint efforts to explore and develop a new retail model. According to Ali, based on big data and Internet technology, the two sides will cooperate in six areas: all-format integration and innovation, new retail technology research and development, supply chain integration, member system interoperability, payment financial interconnection and logistics system coordination.
"The new retail is to use the Internet and big data to ‘ People, goods and fields ’ And other traditional business elements, including the reconstruction of production processes, the reconstruction of the relationship between merchants and consumers, and the reconstruction of consumer experience. In the future, business will no longer be divided into online and offline, and there will be no difference between virtual entities. " Zhang Yong, CEO of Alibaba Group said.
In the field of household appliances, according to the newly released White Paper on Smart Air Conditioning Market in China in 2016, the domestic smart air conditioning industry has been improved in manufacturing, channels, services, resources and application level of new technologies in 2016, especially the introduction and resource sharing of cloud and big data technology, which has laid a solid foundation for resource sharing and service innovation in related industries.
Take Haier, a home appliance company, as an example. Relying on big data, the air conditioners produced by Haier not only built an intelligent energy-saving system, but also applied big data to the fields of heating and power supply. In Jiangsu, Haier’s air-conditioning cloud data has been effectively connected with the State Grid, so as to adjust the peak power load of residents, reduce the peak-valley difference of the grid by 18.47%, and greatly ease the operating pressure of the grid.
Realize "personalized mass production"
In fact, the continuous application of big data is solving the contradiction between "personalization" and "mass production" in the past, which has brought about changes in business models.
In the past, mass-produced products could hardly meet the individual needs of customers. Even with massive data as the marketing basis, we could only subdivide the needs of users as much as possible, and then mass-produce in each subdivision. With the penetration of Internet, big data, cloud planning and other technologies, it is possible to meet personalized and customized needs, and personalized mass production is being tried and explored. According to Chen Feifei, senior director of Vipshop, it is a bold cross-border innovation to apply big data to the fashion industry and conduct e-commerce operation. "Let photo big data, sales big data, artificial intelligence collide with fashion trends, connect data with art, and provide more inspiration and possibilities for creation in the fashion field." Chen Feifei said.
This attempt has brought obvious changes to the production mode of fashion industry, that is, from the former designer-led mode to the user-centered personalized mass production mode. Consumption data will become the most valuable resource in the fashion industry ecological chain. Whoever has more accurate user data will have more market opportunities.
Similarly, in the field of household appliances, Haier air conditioning can simulate 30 body parts, 20 metabolic modes, 162 neuron sensors and 17 warm and cold environments through joint research and development of intelligent bionic human technology. At the same time, relying on bionic people, we have developed original technologies and products such as natural wind, self-cleaning and ion air supply to realize personalized energy saving for thousands of people and make every air conditioner produced in batches "different".
In the field of medical and health care, because of the different disease history and genetic composition of each person, the standardized treatment scheme for diseases in the past is more like a "stopgap measure" under technical restrictions. Nowadays, with the breakthrough of medical sensors, monitors and diagnostic technologies, patient data has become complete and detailed. On this basis, the artificial intelligence system can comb the medical records, genome sequences and other health behavior data of millions of patients, so as to match the personalized treatment plan.
Various trends show that this change in business model is rapidly infiltrating into various industries. The universal landing of "personalized mass production" will also be possible with the support of big data.