In the past decade or so of development, BAT has accumulated massive amounts of data. Baidu has accumulated user search behavior, and this data covers all aspects of life, such as medical care and travel; Alibaba has accumulated data related to user purchasing behavior, from decision-making to purchasing, as well as the goods purchased, and Tencent has accumulated information on user communication.
With the basic data, and then use scientific analysis and processing, in order to produce the value of the user side. If there is no useful data, it is "a clever woman can not cook without rice", and can not be called a real data site, just like you have a secret book of martial arts, but they will not be half a move a style, the same can not be called a martial arts master.
After more than ten years of cultivation in the local O2O industry, Dianping has accumulated more than 42 million evaluation information, more than 10 million merchant information, in addition to 180 million users of mobile data, has perfected the membership system. This is the basis for VWDianping to become a data site and an important milestone. In addition, VWDianping has accumulated a huge amount of data on transactions, user browsing and so on.
The content (i.e., data) on the VWDianping site is generated through UGC, and the data starts to be generated from the first user uploading the first merchant, which currently generates up to a million pieces of content per month. First, the basic information of the merchant is displayed, and then more and more users make reviews, and in the process, the platform accumulates data from both merchants and users.
One is the data about the merchant's address, food, environment, service, etc., and the other is the user's consumption habits, etc., and in the process of the user's review, there is also a mechanism to continuously adjust and improve the merchant's data.
When VWDianping started to involve in the transaction business, the data was even richer. Currently, in VWDianping's big data structure, the user's behavioral log data volume accounts for most of the total data volume, and the rest is transaction data.
Behind the massive amount of user review information includes user preferences for food and drink, the geographic location of activities, and even the transaction information behind the data, and the data is constantly being generated, which meets the basic needs of the socialized division of labor to generate data.
VWDianping is currently launching applications divided into two categories: one is the merchant pass, promotion pass, these are fee-based products; the other category, such as VWDianping index, to provide consumers with decision-making; and catering industry weathervane, to provide reference for the development of the industry; at the same time in the business side of the personalized recommendations can be provided to the user; Dianping housekeeper can help merchants to analyze the business behavior, user characteristics, and so on. At the same time, Dianping is also doing open platform, that is, the data will be open to the third party, they are on this basis for secondary development, towards openness is the trend of big data applications.
Not long ago, the technology department of Dianping made an interesting attempt to refine the information related to the consumption characteristics of the constellation from the massive user review information, and after releasing the information on the Internet, it got a lot of response from the fans. This entertaining data attempt is just the beginning, and more in-depth data exploration can be done later.
In a previous interview, Zhang Tao mentioned that based on the analysis of the user review information in a region, combined with the user click traffic, we can get a lot of information. For example, in a certain city, which cuisine is more popular, which projects are more attention. And Dianping can analyze a region through big data, and even materialize into a business district development level and stage.
The monthly million UGC content generation is just a milestone for Dianping's data, and with the subsequent enrichment of user data, combined with user search, transaction and other data, Dianping can provide users with more and more intelligent products.
Specific to the product is that a person who loves to eat steak, when he is searching for nearby food, the merchants that provide steak will be ranked in the priority position. This is just a simple application, with the richness of the data, the public comment can directly give you advice, attach which merchant's six-minute cooked filet mignon is most in line with your taste. This information is only more intelligent and intimate to the user.
Of course, the whole mystery of the application of big data has only been lifted a little, more in the concept and idea stage, after there is a long way to go.